Skip to main content

Full text of "Enhancements to the Engine Data Interpretation System (EDIS)"

See other formats




y 

f] 



ti 



£.' 



* »= 



lEfc • y lAo^ 4 -i*^ 4 c cw, < -' - 'S 



UAH Research Report 
Date of Issue: July 1993 

Enhancements to the Engine Data Interpretation System (EDIS) 



y Prepared by: 

Martin 0. Hofmann 
|1 Department of Electrical and Computer Engineering 

The University of Alabama in Huntsville 
a Huntsville, AL 35899 

Prepared for: 

George C. Marshall Space Flight Center 
National Aeronautics and Space Administration 
Marshall Space Flight Center, AL 35812 



Final Report on: 

Contract No. NAS8-38609, Delivery Order 35 
Period of Performance: 16 April 1992 to 15 April 1993 



^ Disclaimer Statement: 

tl "The views, opinions, and/or findings contained in this report are those of the 

B authors and should not be construedas an official NASA position, policy, or de- 

cision, unless so designated by other documentation". 



Distribution statement: 

Distribution is unlimited. 



(NASA-CR-192593) ENHANCEMENTS TO N94-10815 

THE ENGINE DATA INTERPRETATION 

SYSTEM (EDIS) Final Report, 16 Apr. 

1992 - 15 Apr. 1993 (Alabama Unclas 

Univ.) 12A p 



G3/37 0180881 



1 



1 



1 



ti 



s 



NASA 



ScdceAorwiaiacr 



Report Documentation Page 



1 . Report No. 



2. Government Accession No. 



3. Recipient s Catalog No. 



4. Titte ano Subtitle 



! 5. Reoort Date 

I 



ENHANCEMENTS TO THE ENGINE DATA INTERPRETATION I 
SYSTEM (EDIS) ! 



July 1993 



6. Performing Organization Code 



7. Authorial 

Martin 0. Hofmann 



i 3. Performing Organization Reoort No. 



10. Work Unit No. 



9. Perforrrung Organization Name and Address ; 

Department of Electrical and Computer Engineering 1 * Contfact or Gfant No - 

; NAS3-38609 D.O. 35 

The University of Alabama in Huntsville : 

HUNTSVILLE, AL 358 99 



12. Sponaonng Agency Name ana Address 

NASA/MSFC 

Propulsion Laboratory 

MSFC, AL 



13. Type of Report and Period Covered 

' Final Report 
4/16/92 - 4/15/93 




The Engine Data Interpretation System (EDIS) expert system project assists the data review personnel at 
NASA/MSFC in performing post-test data analysis and engine diagnosis of the Space Shuttle Main Engine 
(SSME). EDIS uses knowledge of the engine, its components, and simple thermodynamic principles instead 
of, and in addition to, heuristic rules gathered from the engine experts. EDIS reasons in cooperation with 
human experts, following roughly the pattern of logic exhibited by human experts. EDIS concentrates on 
steady-state static faults, such as small leaks, and component degradations, such as pump efficiencies. The 
objective of this contract was to complete the set of engine component models, integrate heuristic rules into 
EDIS, integrate the Power Balance Model into EDIS, and investigate modification of the qualitative reason- 
ing mechanisms to allow "fuzzy" value classification. The result of this contract is an operational version of 
EDIS. EDIS will become a module of the Post-Test Diagnostic System (PTDS) and will, in this context, 
provide system-level diagnostic capabilities which integrate component-specific findings provided by other 
modules. 



17. Key Words <Suggestea by Authonsi) 

Expert System. 

Artificial Intelligence. 

Model-based diagnosis . 

SSME. 

Heuristic Search. 



18. Distribution Statement 



Unclassified-Unlimited 



19. Security Oassif. (of this report) 

UNCLASSIFIED 



20. Security Oassif. (of this pagei 

UNCLASSIFIED 



21. No. of pages 

125 



22. Price 



I 



NASA FORM 1626 OCT 86 



UAH Research Report 
Date of Issue: July 1993 

Enhancements to the Engine Data Interpretation System (EDIS) 



Prepared by: 

Martin O. Hofmann 

Department of Electrical and Computer Engineering 
The University of Alabama in Huntsville 
Huntsville, AL 35899 

Prepared for: 

George C. Marshall Space Flight Center 
National Aeronautics and Space Administration 
Marshall Space Flight Center, AL 35812 

Final Report on: 

Contract No. NAS8-38609, Delivery Order 35 
Period of Performance: 16 April 1992 to 15 April 1993 

Disclaimer Statement: 

"The views, opinions, and/or findings contained in this report are those of the 
authors and should not be construed as an official NASA position, policy, or de- 
cision, unless so designated by other documentation". 

Distribution statement: 

Distribution is unlimited. 



TABLE OF CONTENTS 

LIST OF FIGURES .7. v 

1. Introduction 1 

2. Tasks 2 

3. Component Models 2 

3.1 Component Type PIPE 6 

3.2 Component Type COOLING 8 

3.3 Component Type VALVE 8 

3.4 Component Type PUMP 9 

3.5 Component Type HIPUMP , 10 

3.6 Component Type HYDRAULIC JITJRBINE 11 

3.7 Component Type GASJTURBINE 12 

3.8 Component Type PRE_BURNER 12 

3.9 Component Type MAIN_BURNER 13 

3.10 Component Type CONTROLLER_CONST 14 

3.11 Component Type TWO_SPLIT 14 

3.12 Component Type THREE_SPLIT . . . . 15 

3.13 Component Type UNEVEN JIHREE_SPLIT 15 

3.14 Component Type TWO JOIN . , .- . 15 

3.15 Component Type NOZZLE 16 

3.16 Component Type TANK '. . . 16 

3.17 Component Type SENSOR 16 

4. Diagnostic Reasoning 16 

5. SSME Configuration 22 

6. Heuristic Rules 26 

7. Power Balance Model 28 

8. Fuzzy qualitative system 30 

8.1 Fuzzy qualitative model , 31 

8.1.1 Fuzzy qualitative states 31 

8.1.2 Fuzzy interval arithmetic . . 32 

8.1.3 Fuzzy Constraints 34 

8.2 Implementation 35 

8.3 Comparison against crisp qualitative method 38 

8.4 Limitations of fuzzy qualitative model 39 

8.5 Management of complexity by selective expansion 39 

9. Running EDIS 39 

10. Example 40 

10.1 Standard Operating Mode 40 



iii 



10.2 Using PBM Data 

10.3 Using Heuristic Rules 

10.4 Using Heuristic Rules and PBM Data 



11. Known Limitations 

12. Future Work 

13. Conclusions 



14. References ■. . . 

APPENDIX 

A.1 NEXPERT Code (included only in master copy) 



A.2 SSME Configuration Files 

A.2.1 Component Files 

A.2.2 Fault Mode Likelihoods 



A.3 Heuristic Rules File 



A.4 PBM Data Support Files 



A.5 MCC Leak Example Case Data 

A.5.1 MCC Leak Example: Qualitative values of measured parameters 

A.5.2 MCC Leak Example: Comparison of numerical data 

A.5.3 MCC Leak Example: Standard EDIS execution transcript 

A.5.4 MCC Leak Example: PBM qualitative data file 

A.5. 5 MCC Leak Example: Execution transcript using PBM 

A.5.6 MCC Leak Example: Execution transcript using PBM and heuristic rules 

A.6 Fuzzy Qualitative System Code (included only in master copy) 



41 
41 
41 

42 

42 

43 

45 

46 

47 

262 

262 
282 

283 

289 

302 

302 
303 
317 
323 
326 
332 

335 



iv 



LIST OF FIGURES 

Figure 1: Hierarchy of Structural Component Types 3 

Figure 2: Hierarchy of Behavioral Component Types 4 

Figure 3: SSME Model - Components and Interconnections, Level 1 22 

Figure 4: SSME Model - Components and Interconnections: Block Diagram , 23 

Figure 5: SSME Model - MCC/NOZZLE 23 

Figure 6: SSME Model - COOLING 24 

Figure 7: SSME Model - FUELSUPPLY 24 

Figure 8: SSME Model - LOX SUPPLY 25 

Figure 9: A fuzzy interval . . 32 

Figure 10: Definition of three overlapping fuzzy intervals 32 

Figure 11a: Fuzzy intervals for variable A 33 

Figure lib: Fuzzy intervals for variable B 33 

Figure lie: Fuzzy intervals for variable C 33 

Figure 12a: Object Diagram 35 

Figure 12b: Component object class 36 

Figure 13: Data flow diagram 37 



□ 



Enhancements to the Engine Data Interpretation System 

(EDIS) 



I 



PINAL REPORT 

Enhancements to the Engine Data 

Interpretation System (EDIS) 

NAS8-38609 D.0. 35 



Prepared by 

Martin O. Hofmann 

Department of Electrical and Computer Engineering 

The University of Alabama in Hunts ville 

Huntsville, AL 35899 

April 1993 



1. Introduction 

The Engine Data Interpretation System (EDIS) expert system project was conceived with the goal to assist the 
data review personnel at NASA/MSFC in performing post-test data analysis and engine diagnosis of liquid 
propulsion engines exemplified by the Space Shuttle Main Engine (SSME). EDIS was to use knowledge of 
the engine, its components, and simple thermodynamic principles instead of, or in addition to, heuristic rules 
gathered from the engine experts. EDIS was to reason in cooperation with human experts, following roughly 
the pattern of logic exhibited by human experts. EDIS concentrates on steady-state static faults, such as small 
leaks, and component degradations, such as pump efficiencies, which do not require immediate shutdown or 
similar drastic actions. EDIS systematically analyzes the Behavior of each component of the SSME, search- 
ing for a plausible explanation of the observed^ data anomalies. Triggered by tell-tale anomalies and expert- 
defined fault expectations EDIS hypothesizes a fault and then attempts to prove that this fault is consistent 
with the rest of the data. 

EDIS is not meant to replace review personnel but to facilitate their work. EDIS is capable of providing a 
"second opinion*' that can be contrasted with human data interpretations. EDIS is methodical and will detect 
inconsistencies of a fault hypothesis with the data. It can thus also be used to verify hypotheses proposed by 
review personnel. (The required interface features for this type of behavior have yet to be added though.) 



Final Report July 1993 



Enhancements to the Engine Data Interpretation System 
(EDIS) 

A limited prototype of a knowledge-based post-test data analysis and fault diagnosis system for the space 
shuttle main engine had been constructed under a previous contract. That system demonstrated the validity of 
our qualitative model-based reasoning approach to general engine diagnostic applications. Earlier versions 
of EDIS also performed anomaly detection but the current version expects a set of anomalies as input An 
independent module providedby NASA from a different contract, the PTDS (Post-Test Diagnostic System), 
will provide this data. EDIS will become a module of the PTDS and will, in this context, provide system-lev- 
el diagnostic capabilities which integrate component-specific findings provided by other modules. EDIS 
may be used to reconcile hypotheses generated by specialist modules with the behavior of the engine as a 
whole. 

The objective of this contract was to initiate another phase of development of EDIS, to be used to create a 
complete, useable prototype that will successfully interact with existing numerical models. Four specific 
tasks were identified as listed below. Wfehave successfully addressed all the tasks of the contract as explained 
below. The list of tasks and a short statement of the results of each task appear in the next section. The follow- 
ing sections explain the concepts and implementation of our solutions in more detail. We continue with an 
example, list some known problems, make some suggestions for future work which would enhance the pres- 
ent EDIS system, and give our conclusions. Several appendices contain source code (only available in the 
master copy) and example data. 

2* Tasks 

Task 1: Complete the set of engine component models. Gather NASA MSFC engine systems expertise, 
and apply to the constraint representation using the NEXPERT software tool. 

Results: Completed (see Section 3). Refinements may be needed. 

Task 2: Integrate heuristic rules into EDIS. Subject existing leak rules to critique by NASA engineers. 
Modify the heuristic evaluation function to apply the heuristics. Incorporate the capability for 
the user to enter specific information regarding faults and as well as influence the heuristic evalu- 
ation function. 

Results: Mostly completed (see Section 6). Rules were extracted from interview transcripts prepared by 
engineers from MSFC, LeRC, Aerojet, and Sverdrup. User input must occur through the NEX- 
PERT developer's interface because the Motif-based UIF is not available yet. This is not really 
practical. 

Task 3: Integrate the Power Balance Model into EDIS. Manipulate data to be accessible by NEXPERT. 

Results: Mostly completed (see Section 7). Data has to be transferred and formatted manually. 

Task 4: Investigate modification of qualitative reasoning mechanisms to allow uncertainty for value 
classification. Use fuzzy logic to describe uncertainty. 

Results: A methodology was developed (see Section 8) but no complete diagnostic system has been coded 
yet. 

3. Component Models 



The EDIS system contains a collection of basic thermodynamic components from which arbitrary systems 
can be configured. The configuration is read by EDIS at the start of processing from a specified sub-directory. 



Final Report, July 1993 



u 



Enhancements to the Engine Data Interpretation System 

(EDIS) 



Each component type is associated with a specific file. The file names for the components are listed with the 
description of the component models which follows. The SSME configuration is described in detail in sec- 
tion 5 

Component models have two major parts; one part, the structural part, describes the interconnections among 
components, and another part, the behavioral part, specifies the ways in which a component may be behaving. 
In addition, faults which affect component types are described in terms of deviant behavior and their relative 
likelihoods are specified. Structural and behavioral model templates are each organized in a class hierarchy. 
Figures 1 and 2 show the two hierarchies. The behavioral type hierarchy is very shallow because the behavior 



COMPONENTS 
STATIC COMPONENTS 

- v \ " 

MANIFSId CONTROLLER SENSOR THERMO COMPONENT TANK BURNER 




/\ \ 

tJPLIT 1 CONTROL 



// \ 

Jnozzle^val 



I 



TWOJPLIT 1 CONTROLLERJCONST ^NOZZLE 1 VALVE MAIN_BURNER 

I TWO JOIN DUCT ENERGY_CONV_COMP PRE_BURNER 



THREE SPLIT COOLING PIPE 



// I \ 



PUMP TURBINE 



ivi 



I \ 



UNEVEN THREE SPLIT 



HIPUMP GASJVRBINE HYDRAULIC [TURBINE 



Figure 1: Hierarchy of Structural Component Types. Abstract types 
are shown in roman font, types for which instances are de- 
fined are shown in bold italics. 



of each type is represented using type-specific rules and no inheritance mechanism exists among rules. Most 
behavior rules, however, make use of common building blocks to describe component behavior. These build- 
ing blocks are derived from the constraint types used In defining component behavior, see below. 

The behavioral model associates with each component a set of qualitative-valued parameters whose values 
represent the momentary behavior of the component. Normal component behavior is characterized by a set of 
constraints associated with the component. For steady-state analysis these constraints can be derived from 
conservation laws. For example, a pipe in a thermodynamic system carrying fluid is characterized by the dif- 
ference of the energies of the fluid mtenng lEd leaving friction, and a pump transforms 
mechanical energy supplied to its shaft into fluid energy. Commonly, parameters representing energy cannot 
be measured directly but are derived from constituent parameters. For example, fluid energy depends on pres- 



Final Report July 1993 



Enhancements to the Engine Data Interpretation System 
(EDIS) 



COMPONENT 



TANK 
NOZZLE 
TWOJOIN 
TWO_SPLlT 
PRE_BURNER 
MAIN_BURNER 
CONTROLLER_CONST 

THREE SPLIT 




UNEVEN 



£- 



TWO PORT 

N 

PIPE 
VALVE 
COOLING 
dASTURBINE 
HYDRAULIC TURBINE 



SPLIT 



PUMP 

\ 

HIPUMP 



Note: Every class shown by its name "NAME", e.g. PIPE, in this graph is actually called 
"NAME"_BEHAVIOR, e.g. PIPE_BEHAVIOR, in the program code. The suffix 
"_BEHAVIOR" is omitted to enhance readability. 

Figure 2: Hierarchy of Behavioral Component Types 



sure, height, and flow rate. In steady-state, only deviations from normal values are of interest. Therefore, the 
parameter values may be restricted to the qualitative values NORMAL, LOW, HIGH, and the special label 
Unknown. Also, the conservation equations may be simplified, e.g. by linearizing, or may even be trans- 
formed into qualitative confluences [1]. 

Parameters associated with a component fall into one of two categories, measurable and derived. Derived 
parameters are related to measurable parameters through relations which do not depend on the state of the 
component, called "mathematical constraints" (M.C). Relations between parameters representing energies 
and other conserved quantities are called "fundamental constraints" (EC.) and characterize component be- 
havior modes. After simplification, linearization, and transformation into the qualitative domain, constraint 
expressions are called "incremental qualitative constraints" (IQCs). 

It has been shown by Kalagnanam et al. [7] that the ordinal properties of the involved quantities do not change 
even under such strong simplifications as long as the simplifying transformations are monotonia Our simpli- 
fications and transformations from quantitative to qualitative models therefore preserve relative magnitude of 
parameter values. If, for example, the qualitative model predicts an increase in value then the quantitative 
model (if it existed) would also predict an increase. Invariance of ordinal properties in essence guarantees that 
qualitative values are predicted correctly by IQC's. 

Five types oflQCs are defined. The two-place relation "proportional" (p/2), the three-place relations 
"qualitative-synergy" (®/3), "qualitative-antagonism" (0/3), and "qualitative-optimum" (0/3), and the 
four-place qualitative-synergy (0/4), an extension of the three-place synergy. These relations are best de- 
fined using relation tables. The four-place qualitative synergy is not listed because it can be derived from its 
three-place version. 



Final Report, July 1993 



V 



Enhancements to the Engine Data Interpretation System 

(EDIS) 



© (A,B,C,D) = ((A^C.,!),) | 3x : © {x,B„C^ A © (A b x,D^, x E {LOW,NORMAL,HIGH}} (eq. 1) 

The NEXPERT implementation uses explicit representations for three- and Four-place relations; they are 
stored as sets of tuples in files gtypei.nxp (®/3), gtypeiLnxp (0/3), gtypeiiLnxp (®/4), and gtypeiv.nxp 
(Q/3). In these files parameters have generic names A, B, C, and D. Three-place relations are to be read as 
0(A,B,C) or equivalently (B O C) -* A, and four-place relations as 0(A,B,C,D) m(BOCOD)-»A 

Relation p 



Relation 0/3 



Relatione/3 



P 



LOW 



NORMAL 



HIGH 



LOW 



NORMAL 



HIGH 



B0C—A 


C 


B 


LOW 


NORMAL 


HIGH 


LOW 


LOW 


LOW 


III 


NORMAL 


LOW 


NORMAL 


HIGH 


HIGH 


LOW, 

NORMAL, 

HIGH 


HIGH 


HIGH 



m 



w 



B0C-A 


C 


B 


LOW 


NORMAL 


HIGH 


LOW 


LOW, 

NORMAL, 

HIGH 


LOW 


LOW 


NORMAL 


HIGH 


NORMAL 


LOW 


HIGH 


HIGH 


HIGH 


III 



Relation 0/3 



Final Report July 1993 



Enhancements to the Engine Data Interpretation System 
(EDIS) 



B0C—A 


C 


B 


LOW 


NORMAL 


HIGH 


LOW 


NORMAL 


LOW 


LOW 


NORMAL 


LOW 


NORMAL 


LOW 


HIGH 


LOW 


LOW 


NORMAL 



For example, the pressure difference in a pipe is derived from input and output pressures via a mathematical 
constraint. Fluid velocity in the pipe determines energy loss due to friction. The energy balance between 
input and output of the pipe is expressed by a balance between fluid velocity V and pressure difference p^. 
This balance is captured by the fundamehM constraint V - p^. "The fact that the pipe does not normally 
leak (pipe branches are represented separately) is expressed by the F.C. Inflow = Outflow, derived from mass 
conservation. 



Behaviors of connected components are interrelated through the parameters shared at the interface between 
components. For example, if a pipe is connected to a pump, fluid pressure, temperature, and velocity at the 
pipe output are identical to fluid pressure, temperature, and velocity at the pump input. Note that although in 
fluid systems inputs and outputs can be distinguished, the constraint model is non-directional. Connection 
constraints are simply equality constraints on the qualitative values. In the following sections we will devel- 
op models of all the component types implemented in EDIS. 

One of the important features of EDIS is that the models can express that certain behaviors are physically 
impossible under reasonable assumptions. For example, no heat or energy is transferred to the fluid or gas 
except where explicitly specified. No mass can be introduced into the system except from the tanks. These 
additional "physical constraints" help in reducing the number of assumptions which may realistically be 
made about component and system behavior. 

The heuristic evaluation function implemented in EDIS matches component behaviors against fault modes 
specific to each component. If the behavior matches a fault mode, its likelihood is adjusted to reflect the 
likelihood of the matching fault mode. We have defined a small set of fault modes and we will note in the 
discussion of each component type what fault modes are currently being tested for. 

3.1 Component Type PIPE 

The behavior of a pipe is characterized by energy conservation, see equations (2) and (3), and a mass con- 
servation equation (4) between the pipe inlet and the pipe outlet. We assume that any possible changes in inlet 
and outlet temperatures are irrelevant to the diagnosis, see equation (5). A separate model for cooling ducts 
models temperature changes caused by heat transfer. 



V? 

in 

2F 



+ z i- + 



-^ _ F - Yk + 

Y c Loss - 2g 



Pout 



'•out 



E - f V L 

E Loss - l 2g D 

Yin "in in = Yout ^out " out 

Qi„ = Qout 



(eq. 2) 

(eq.3) 

(eq.4) 
(eq.5) 



Final Report, July 1993 



\j Enhancements to the Engine Data Interpretation System 

(EDIS) 

w V = average fluid velocity 



fm 



V 


= fluid velocity 


g 


= gravitational constant 


z 


= height 


P 


= pressure 


Y 


= density 


L 


= pipe length 


D 


= pipe diameter 


f 


= friction coefficient 


A 


= pipe cross-sectional area 



Q = heat flow rate 

After linearization and simplification equations (2) through (5), reduce to (6) through (8) respectively. The 
delta operator (A) indicates incremental (small signal) change and K is a constant which depends on the oper- 
ating point, the pipe dimensions, and the friction coefficient. 

A (Pin " Pout) = K • AV (eq>6) 

AV in = AV out (cq.7) 

AT ia = AT o«. (eq.8) 
T = temperature 

The essence of these equations, which is captured by IQCs, is that the pressure difference between inlet and 
outlet is proportional to the velocity, and that the input velocity is proportional to the output velocity as long as 
the pipe is operating correctly. Also, Temperature changes at the input are passed through the pipe unchanged. 
Faults which could invalidate the constraints are pipe leaks and obstructions, for example. Finally, we can 
formulate the IQCs. Note that all parameters in the IQCs represent small changes from an operating point. 

Fundamental Constraints: 

Pdiff P V («1- 9 ) 

Vj, p V out (cq.10) 

Mathematical Constraints: 

Pin 9 Pout -* Pdiff («1- n ) 

V in 0V out -*V (eq.12) 

Assumptions: 
In addition, some simplifying assumptions are being made. 

T* « T out (eq. 13) 

Physical constraints: 



M Final Report My 1993 



Enhancements to the Engine Data Interpretation System 
(EDIS) 



V ut 


£ 


v* 




Fault modes: 








Leak: 






V ou t < V^ 


Obstruction: 






Pdiff > V 


3.2 Component Type COOLING 









(eq. 14) 



A cooling duct behaves like a pipe except that heat is transferred into the medium (fuel in our case) from the 
cooled component. We made a major simplification and assumed that the temperature T source of the cooled 
component is not changed by changes in the temperature and flow of medium in the cooling duct. Feedback 
from the cooling duct to the cooled component is not modeled. All constraints, assumptions, and fault modes 
of component type PIPE apply and the following constraints are added. Changes in temperature increase 

from cooling duct input to cooling duct output are positive for increases in heat inflow (Qi n ) and negative for 
increases in mass flow rate through the cooling duct (V) (eq. 15). Heat inflow is determined by the tempera- 
ture difference between T sourcc and Ti n (eq. 16). 

Mathematical Constraints: 

QiaGV-T^ (eq.15) 

Tsourcc 9T in -Q in (eq.16) 

ToutOT^T^ (eq.17) 

In the implementation we neglect changes in cooling duct input temperature (Ti n ) because T source is much 
larger than Tj n . Then, changes in heat inflow are equivalent to changes in heat source temperature (T sourcc ) 
and(^l^)and(eq. 16) simplity to(eq. 18). 

Tsource G V -* T diff (eq. 18) 

3,3 Component Type VALVE 

The model for component type VALVE is similar to the PUMP model except that the pressure difference be- 
tween input and output now also depends on valve position. We assume that temperature does not change 
between input and output and we do not allow for leaks in a valve, i.e. input and output massflow rate are 
identical. We model the translation of the valve position command into the actual measured position by 
(eq. 20). Valve failure can occur if the valve is blocked, for example, and (eq. 19) is violated, or if the valve 
does not respond correctly to the position command from the controller and (eq. 20) is violated. Next, we list 
the complete set of constraints for type VALVE. 

Fundamental Constraints: 

Pdia P V0 position (eq. 19) 

position p commandedjx)sition (eq. 20) 

Mathematical Constraints: 

Pm e Pout -* Pdiff («i- 2i) 



Final Report, July 1993 



Enhancements to the Engine Data Interpretation System 

(EDIS) 



Assumptions: 



Physical constraints: 



Fault modes: 



T = T 
x in * out 

^m = * out ~ V 



p di£f ^ V9 position 



(eq.22) 
(eq.23) 

(eq. 24) 



Blockage: 
Servo fault: 



Pdiff > deposition 
position * commanded_position 



3,4 Component Type PUMP 

To analyze a pump we again start with the energy balance equation (eq. 25), i.e. the first law of thermodynam- 
ics, this time written as a rate equation for a steady-state, steady-flow process. We neglect potential energy. 



(eq.25) 





W + Q + m ia 


f V 2 1 


" *out 


r V 2 ' 

out 
PoutV + 2 


Q 


= heat transfer rate 


xh 


= mass flow rate 


W 


= incoming power 


V 


= fluid velocity 


V 


= specific volume 


The mass balance demands 






^in = *oi 


It = m 





(eq.26) 
Assuming an adiabatic process where Q = and letting V in = V out gives 

W = m(p out - Pin )v (cq-2 7) 

Next we can replace mv by AV, where A is the pipe cross-sectional area and V is fluid velocity, and get 

W = (Port-pujAV (eq-28) 

We now introduce the qualitative variables MechPWR and PV_Product which stand for the expressions on 
either side of equation (eq. 28). In addition to equation (eq. 28) we have a relation between input and output 
mass flow rates from equation (eq. 26) and a relation between the rotational speed of the pump (co) and the 
effecti ve velocity of the fluid through the pump (V) (eq. 29). An analysis of several data sets collected during 
test firings of the SSME shows, however, that this relation does not hold for the low pressure pumps. We must 
assume that turbulence and seal leakage have a large effect on this relation. It is therefore ignored for type 
PUMP but enforced for type HBPUMP which is used to represent the behavior of the high pressure pumps. 



Final Report July 1993 



Enhancements to the Engine Data Interpretation System 
(EDIS) 

Also, we are not modeling leak faults in a pump. Leaks are only considered in pipes and cooling ducts. There- 
fore, the mass balance which was a fundamental constraint for type PIPE now becomes an assumption. 
Again, we ignore changes in temperature within the pump. Nevertheless, it may be advantageous to describe 
pump inefficiencies by the temperature rise they cause in the fluid being pumped. Pump efficiency faults may 
be easier to represent and find using such an extended model. We are considering this enhancement for the 
future. 

Mechanical power (MechPWR) can be derived from shaft speed (o) and torque (Tq) (eq. 32), the pressure 
difference (p d iff) as before from p in and p out (eq. 34), and PVProduct from fluid velocity (V) and pressure 
difference (p d iff) (eq. 33). In a pump, pressure difference is calculated as p ut-Pin, a positive quantity. The 
behavioral model for a pump can now be formulated. 

Fundamental Constraints; 



MechPWR p PV_Product 






(eq.30) 


V p (D 






(eq.31) 


Mathematical Constraints: 








Tq ® 0) — MechPWR 






(eq.32) 


Pdiff ® ^ -* PV_Product 






(eq.33) 


Pout Q Pin ~* Pdiff 






(eq.34) 


v ta ©v out -v 






(eq.35) 


Assumptions: 








M n = "out 






(eq.36) 


Tjn = -1 out 






(eq.37) 


Physical constraints: 








MechPWR g PV_Product 






(eq.38) 


(o & V 






(eq.39) 


Fault modes: 








Impeller Problem: 


(0 


> V 




Low Efficiency: MechPWR 


> PV_Product 




3.5 Component Type HIPUMP, 









The type HIPUMP models the behavior of the high-pressure pumps used in the SSME. The high-pressure 
pumps produce an extremely large increase in pressure level from input to output. The qualitative model 
manipulates relative changes in parameter values and therefore has to be careful to interpret deviations with 
respect to the appropriate steady-state level. Pressure deviations at the input have two different reference 
levels, the low pressure-level of the upstream components and the high-pressure level of the downstream 
side. The pressure at the high-pressure pump input therefore may have two possible diverging interpreta- 



Final Report, July 1993 



Enhancements to the Engine Data Interpretation System 

(EDIS) 



tions, i.e. qualitative values. The parameter associated with the high-pressure pump holds a value which 
corresponds to the high-pressure level. The output pressure parameter of the upstream component holds the 
qualitative value with respect to the low-pressure level. In order to maintain these two interpretations the 
following changes were made to the PUMP model. 

• Deviations at pi n which were derived on the low pressure (upstream) side are ignored by the high-pressure 
pump. 

• If analysis of the high-pressure pump predicts pj n to be NORMAL, this value is not propagated to the 
upstream low-pressure component because a much finer scale is used there to detect anomalies. 

• When analyzing the behavior of a high-pressure pump pi n is never assumed to be LOW because such a 
deviation is either small enough to be neglected or serious enough to interfere with correct functioning of 
the SSME. EDIS only deals with "small" anomalies. 

• If input pressure is measured the measured value is interpreted from the low-pressure side point of view. 
In addition, another assumption holds for the high pressure pump type. 



Additional Assumption: 



V = co 



(eq. 40) 



3,6 Component Type HYDRAULICJTURBINE 

The behavioral model of type HYDRAULICJTURBINE is identical to the PUMP model except that the pres- 
sure difference is taken from input to output. Also, the inequalities which characterize fault modes and physi- 
cal constraints are inverted because mechanical power now leaves the component. 



Fundamental Constraints: 



MechPWR p PV_Product 

V p a) 



Mathematical Constraints: 



Assumptions: 



Tq © (0 -* MechPWR 
Pdiff © v -* PV_Product 

Pin © Pout -* Pdiff 
^in © "oul ~* V 





in ™" out 




* io " * out 


cal constraints: 






MechPWR g PVProduct 




0) ^ V 


modes: 





(eq.41) 
(eq.42) 

(eq.43) 
(eq.44) 
(eq.45) 
(eq.46) 

(eq.47) 
(eq.48) 

(eq. 49) 
(eq. 50) 



Final Report July 1993 



11 



Enhancements to the Engine Data Interpretation System 
(EDIS) 



Impeller Problem: co < V 

Low Efficiency: MechPWR < PV_Product 

3*7 Component T>pe GASJTURBINE 

In a gas turbine the first law of thermodynamics equates mechanical power produced (MechPWR) to the dif- 
ference in enthalpy of the gas entering and leaving the turbine (h^ff) (eq. 51). We are neglecting differences in 
gas velocity and assuming an ideal gas. 

Fundamental Constraints: 

MechPWR p h m (eq. 51) 

hdiff= difference in enthalpy of entering and exhausted gas 
Mathematical Constraints: 

K © hout ~* MechPWR (eq. 52) 

hj n = enthalpy of gas entering the turbine 
ho Ut = enthalpy of gas leaving the turbine 
Assumptions: 

V in = V 0tt( = V (eq.53) 

K = Pm = T ln (eq. 54) 

hour = Pout = T out (eq. 55) 

MechPWR £ h diff (eq. 56) 



Physical constraints: 



Fault modes: 



Low Efficiency: MechPWR < h 



diff 



3,8 Component Type PREBURNER 

A pre-burner produces a fuel-rich hot gas through incomplete combustion which drives a high-pressure 
turbo-pump and which eventually reaches the main combustion chamber where it is burned completely. The 
equations which govern the combustion process are once again derived from energy balance equations. The 
enthalpy created in the incomplete combustion process, i.e. the enthalpy of formation of the steam produced, 
can be simplified to a linear function of the mixture ration (MR). Enthalpy is determined from temperature 
under ideal gas conditions. The complete enthalpy balance equates enthalpy of the products (ho Ut ) to the prod- 
uct of mixture ratio (MR) and mean inflow temperature (T). We linearize and simplify this product relation 
and rewrite it as an incremental qualitative-synergy of mixture ratio (MR) and mean inflow temperature (T) 
(eq. 57). Mass conservation equates input flows (V ox , Vfu C i) with output flow (V out ) (eq. 58). 

Fundamental Constraints: 

h out p MR T (eq. 57) 

hout = output enthalpy of partially burned fuel 



12 Final Report, July 1993 



Enhancements to the Engine Data Interpretation System 

(EDIS) 

MR= mixture ratio (oxygen vs. fuel) 
T = mean input temperature of oxygen and fuel 
Mathematical Constraints: 

Vox©V fucl -V out (eq.58) 

V ox eV fuel -MR (eq.59) 

V ox = oxygen input mass flow rate 

VfueF fuel input mass flow rate 

T ox®T fueI ->T (eq.60) 

Tox = oxygen input temperature 

TfueF fuel input temperature 

We observe that the output pressure produced by the preburner back-pressures the fuel input. We assume that 
changes in p out are translated directly into changes in pj n since the pressure produced by the preburner is much 
higher than the fuel input pressure. 

Pout P Pin " (eq-61) 

Pi n = fuel input pressure 

In the implementation, we assume that the controller will keep the fuel flow fairly constant and regulate oxy- 
gen flow to the preburners via the preburner oxygen valves. Thus the incremental quantity Vfuci is zero and 
MR, i.e. changes in the mixture ratio, can be equated to V ox , i.e. changes in oxygen input flow. 

Simplification of (eq.59) V ox p MR (eq. 62) 

Assumptions: 

h ut = Pout = T out (eq. 63) 



Physical constraints: 



Fault modes: 



h om ^ MR 8 T (eq. 64) 



No fault modes are defined yet for the preburner. If lower than anticipated output enthalpy (hout) was observed 
due to some problem with the combustion process itself then this behavior could be defined as a fault mode. 
Mixture ratio problems, however, are external to the preburner, 

3,9 Component Type MAINJBURNER 

We modeled the main burner as if it were operating at optimal mixture ratio. Therefore our model predicts that 
any change to higher or lower mixture ratio will lead to lower engine output. This assumption appears to be 
wrong since the controller operation indicates that power still increases with oxygen flow and therefore with 
higher mixture ratios. Power also depends on the total amount of fuel and oxygen supplied to the main burner. 
Power is equated with output pressure, enthalpy, and temperature. The MAIN_BURNER type also has provi- 
sions to attach a cooling component to it. 



Final Report July 1993 13 



Enhancements to the Engine Data Interpretation System 
(EDIS) 

Fundamental Constraints: 

Pout P V baiancc © V ou , (eq. 65) 

v baiance=a quantity which represents the optimum mixture ratio 
Mathematical Constraints: 

VoxOV fuel -V balancc (eq.66) 

VoxeV^-V^ (eq.67) 



Assumptions: 
Physical constraints: 
Fault modes: 



hout = Pout = T om (eq. 68) 

p om ^ V balance V out (eq. 69) 



No fault modes are defined yet for the main burner. If lower than anticipated output power (p ou t) was observed 
due to some problem with the combustion process itself then this behavior could be defined as a fault mode. 

3.10 Component T^pe CO^^^lOIXERJ:ONST — 

Type CONTROLLER_CONST models a controller which is supposed to keep a parameter value at a constant 
level, i.e. the parameter should have value NORMAL, by setting a control input parameter value appropriate- 
ly. Such as controller is considered to be operating normally as long as the controlled parameter has value 
NORMAL. The CONTROLLER_CONST model has been specifically designed to model the fuel flow con- 
trol system of the SSME. It is the least generic component of the system because input and output parameter 
names have to be defined in our models. Type CONTROLLER_CONST measures a parameter named V in 
and controls a component (usually a valve) through a parameter named "commanded_position." 

Fundamental Constraints: 

V^ = NORMAL (eq. 70) 

Fault modes: 

Controller fault: V^ * NORMAL 

3.11 Component Type TWO_SPLIT 

Component type TWO_SPUT models a pipe "T" with one input and two outputs. It does not include any 
straight pipe sections. Wc therefore assume that the pressures at all its terminals are equal, that the tempera- 
tures are equal and that the sum of outflows is equal to the inflow. No faults are associated with pipe splits and 
joins. Output ports are distinguished by labels A and B. 

Mathematical Constraints: 

Vou^eV^B-V^ (eq.71) 

V out _ A =output flow rate into port A 
V ut B= out P u t flow rate into port B 



14 Final Report, July 1993 



w 



Enhancements to the Engine Data Interpretation System 

(EDIS) 

Assumptions: 

T in = T out_A = T 0Ut _ B (eq. 72) 

Pin = Pout_A = Pout__B (eq. 73) 

3,12 Component Type THREE JSPLIT 

Component type THREE_SPLFT is an extension of type TWO_SPLTT for the case of one input and three 
outputs. This type is not truly necessary and a component of type THREE_SPLTT could be replaced by a 
sequence of two components of type TWO_SPUT. This type has been added for convenience, however 
Output ports are distinguished by labels A, B, and G 

Mathematical Constraints: 

V ou L a © V ou L b © V om.c "* v in (eq. 74) 



Assumptions: 



* in * Aiit A *■ mil "R ^ i 



om_A ~ l out_B ~ l out_c (eq. 75) 

Pin ~ PouA = PoutJB = Pout_C (eq. 76) 

3.13 Component Type UNEVEN JTHREEjSPLIT 

The type UNEVEN_THREE_SPLTT was created as modification of THREE_SPLIT to address the case 
where one branch of the outflow is significantly smaller than the other two. Similar to the problems addressed 
by component type HIPUMP, the different operating levels make it hard to classify deviations consistently 
from the points of view of large and small normal flow rate. An example for this situation is found in the 
DIFFUSER where the amount of fuel flow to the MCC cooling duct is much smaller than both the flow into 
the nozzle cooling and the CCV valve. 

The model basically ignores the small outflow into port C. No value is assigned to the flow parameter V out c 
unless it can be determined from the value set at the MCC cooling duct input. 

Mathematical Constraints: 

V 0Ut _ A V ^ B - Vfc (eq. 77) 

Assumptions: 

Tin = T out_A = T out _B = T out _ c (eq. 78) 

Pin = Pout_A = Pout_B " Poui_C (eq. 79) 

3.14 Component Type TWO JOIN 

Component type TWO_JOIN models the joining of two input flows into a single output flow. Again we as- 
sume that the pressures are forced to be equal but we derive the output temperature from the magnitudes and 
temperatures of the input flows. 

Mathematical Constraints: 

v injv ® V injB -* Vo«t (eq. 80) 

Qin_A ® Qin_B "^ Qout (eq. 81) 



Final Report July 1993 15 



Enhancements to the Engine Data Interpretation System 
(EDIS) 



T in _ A © V inA - Q inA (eq. 82) 

T in _B ® V in_B - Qin.B («!• »3) 

Assumptions: 

Pin_A = Pin.B = Pout (eq. 84) 

3.15 Component Type NOZZLE 

The type NOZZLE has no behavior constraints associated with it. Output temperature is the only parameter 
of interest because it has a cooling component associated with it. Deviations in output temperature are as- 
sumed to directly follow deviations in input temperature. 

Assumptions: 

T* = T out (eq. 85) 

3.16 Component Type TANK 

No constraints are defined for type TANK. It is defined in order to provide boundary components to the 
SSME model. 

3.17 Component TVpe|^[50R 

Type SENSOR is neither defined nor used at this time. If defined, sensor faults could be included in the fault 
diagnosis. Diagnosis becomes less efficient with larger numbers of components, however, and a separate 
module will address sensor faults. 

4. Diagnostic Reasoning 



Diagnostic reasoning is realized by a heuristic A* search methodology. Component are analyzed one by one 
until the behavior of the SSME is completely determined. When a component is analyzed all its possible 
behaviors are enumerated. Each component behavior is rated according to the estimated likelihood that it 
represents the actual behavior of the component. Each new behavior is combined with the behaviors already 
analyzed and global likelihoods for the resulting behavior hypotheses are calculated. A set of component 
behaviors is called a "scenario. " Scenarios created early in the search contain behaviors for only a few com- 
ponents. After the last component has been analyzed, scenarios exist which contain completely specified 
behaviors for all components of the SSME. 

EDIS operates on a single scenario at a time. Whenever a component is analyzed and its behaviors are gener- 
ated and attached to the current scenario, multiple successor scenarios are generated. The heuristic evaluation 
function identifies the most likely among them and this most likely scenario is chosen for further processing. 
Several choices in this process are critical for the performance of EDIS: the order in which components are 
chosen for analysis, the "local" evaluation of each new component behavior, and the global evaluation of the 
scenario made up of a number or scenarios. 

More formally, we define a behavior of the device D to be diagnosed, which is composed of a set of compo- 
nents Comp, as a set of parameters P and a function Beh : P -* Vol which assigns each parameter inP one of 
the elements in the set of qualitative values Val. The set Val is currently defined as Vol = {NORMAL, HIGH, 
LOW, Unknown}. A behavior description is called a scenario. In addition, a scenario contains a function 



Final Report, July 1993 



^^ 



Enhancements to the Engine Data Interpretation System 

(EDIS) 

Mode : Comp -* BM which maps each constituent component in Comp into a behavior mode in BM or the 
special symbol ToBeAnalyzed if the behavior mode remains to be determined. One possible behavior mode 
of each component is the NonFauIty mode; fault modes are defined individually for each component type. 

A finished scenario is a scenario whose Mode function maps every component in Comp into BM - ToBeA- 
nalyzed and whose function Beh maps every parameter in? into a value in Val - Unknown. A partial sce- 
nario is a scenario whose Mode function maps at least one component in Comp into ToBeAnalyzed. A 
finished scenario represents a solution which identifies the faulty components), i.e. all those components 
which are mapped into something other than NonFauIty, and explains in detail how the behavior of the device 
has changed because of the fault(s). By itself , a shift in the values of the parameters associated with a compo- 
nent does not necessarily imply a fault of this component; it can be caused by a shift in operating point due to 
changed input or output conditions. 

Diagnostic search progresses by means of execution of search operators. Search operators expand a partial 
scenario 5 and generate its successor scenarios. Partial scenarios without successor nodes are active or open. 
Operators determine additional parameter values, i.e. they change the image of a subset of device parameters 
under the mapping function Beh from Unknown to values in Val - Unknown. A particular operator must be 
defined for each component type. However, operators are composed, in part, of generic expansion functions 
which apply to qualitative confluences, such as "+" and "-" as they appear in the qualitative constraints 
which define correct component behavior. In our implementation the consistent value assignments for each 
qualitative confluence are pre-computed and cached as a value tuple list. Thus a form of the arc consistency 
algorithm [4] is applied to a subset of the nodes in the constraint network, increasing the efficiency of the 
algorithm. Enumeration is accomplished by selecting only those entries from the value tuple list which con- 
form to the parameter values already chosen. 

The set of constraints associated with a component is analyzed in such a way as to minimize guessing. 
Constraints which operate on larger numbers of parameters and those which define fewer legal value tuples 
are satisfied first. These strategies conform to the "constraint arity" and "constraint tightness" heuristics in 

PI- 



Fundamental constraints are ignored, i.e. suspended, by expansion operators but mathematical constraints 
— and inviolable physical conservation constraints are enforced. When the behavior of a pipe is expanded, for 

|g example, the mathematical constraint p^ = p m - />«* is enforced and restricts the value combinations which 

may be assigned to p £ & /\,,and p^. The inviolable constraint V* ^ V^ restricts the possible value assign- 
= ments to the input and output velocity parameters V m and J^. The expansion operation accomplishes what 

W normally constitutes the first step in a qualitative simulation [1]; it determines the possible initial state of the 

device. Here, no dynamic behaviors are considered and thus no additional qualitative simulation mechanisms 
5 are needed. 7 Zl _ 

Successor scenarios enumerate all possible behaviors of the component whose behavior was expanded last, in 
s the context of what was already known about the component behavior from measurements and previously 

~ made assumptions. The set of possible behaviors is, in general, much smaller than the unrestricted set of be- 

haviors implicitly implied by straight-forward constraint suspension. The proposed algorithm thus develops 
m a more detailed diagnosis than constraint suspension or Kieger's algorithm but, on the other hand, has to rep- 

" resent behavior explicitly which is less efficient. Ojperators can exhaustively expand behavior because only a 

_ small subset of parameters is assigned values at a time and because parameters associated with a component 



Final Report July 1993 17 



Enhancements to the Engine Data Interpretation System 
(EDJS) 

are tightly coupled, by mathematical constraints. The search space is thus hierarchically decomposed even 
though the device description is not necessarily hierarchical. 

Consistency of parameter assignments within one component is guaranteed by the expansion operators. 
Global consistency of assignments within a scenario is guaranteed because each scenario maps each parame- 
ter into a single value. Different components which share one or more parameters must therefore agree on 
their values. If an operator attempts to change the value of a parameter which is anything other than Un- 
known, then this particular successor scenario being developed becomes invalid and is removed. 

Each active partial scenario constitutes a node in the search tree competing to be expanded. A heuristic evalu- 
ation function ranks the active partial scenarios and selects the best one for further expansion. The evaluation 
function judges the parameter assignments already made and estimates the change in cost which is likely to 
accumulate until the scenario is fully expanded. According to the standard definition of the A* algorithm [5, 
page 76], the heuristic evaluation function / is calculated as / = g + h T where the quality of the expansion 
achieved so far corresponds to the cost g of the path from the initial node to the current node, and the expected 
worsening corresponds to the expected cost h! of the remaining path. 

The function g which judges the quality of a particular scenario takes jntp account the merit of each identified 
behavior mode and the number of components yet to be analyzed. The cost estimate h' depends on whether a 
fault has been hypothesized yet and on the results of expanding similar partial scenarios. The cost function g 
is parameterized by a set of merit figures assigned to each behavior mode by a domain expert, which may be 
modified for a particular application, if necessary. 

The general evaluation algorithm defines the cost function g as 

* U* " (eq-86) 

where q ; = q{Mode{comp$) 

and the function q % maps each behavior mode of component comp g into a figure of merit supplied by the do- 
main expert, and ^(ToBeAnalyzed) = 1. The product is taken over the behavior modes of all n = \Comp\ 
components ofthedevice. The figure of merit for each behavior mode is ^ 1 so that the combined cost of two 
or more fault modes is larger, in general, than that of a single fault. The figure of merit of the NonFauIty 
behavior is usually defined as 1. The product rule is motivated by the assumption that faults are independent 
and the fact that the joint probability of independent events is given by the product of their individual proba- 
bilities. A more sophisticated evaluation function could take joint probabilities of interrelated failure modes 
into account. The ratio njn is the ratio of the number of yet to be analyzed components over the total number 
of components. It slowly decreases as more components are analyzed. It is included to keep the line of rea- 
soning from skipping between different branches in the search tree, i.e. to favor depth-first processing, which 
facilitates cooperation with a human user. 



Behaviors whose mode is normal are ^subjected to another test in order to identify more and less likely ones 
among them. The quality of each normal behavior is reduced acoording to the following procedure. For each 
behavior the number of parameters is recorded whose values is not NO RM AL. Behaviors with the lowest 
number of non-normal parameters are considered best and their quality ratings remain unchanged. The quali- 
ty of all other behaviors is reduced according to how many more non-normal parameters they include than the 



lg Final Report, July 1993 



V 



Enhancements to the Engine Data Interpretation System 

(EDIS) 

best behaviors. The difference in the number of non-normal parameters is multiplied by 0.01 to derive the 
penalty for each behavior. 

The expected cost estimate V is defined as 

K = h x ' + h 2 (eq. 87) 

if no fault has been hypothesized yet 



v - p 



1 * * otherwise 

A/ anticipates that at least one fault will be found and ensures that promising failure modes are considered 
early on. A 2 ' is adjusted dynamically as information about global consistency becomes available. If a fault is 
hypothesized but the scenario has to be abandoned later because its global quality becomes too low, h 2 f is set 
to a value which measures the observed worsening of scenario quality. This value of V is applied to all sce- 
narios derived under this fault hypothesis. 

Another use of h z f would be in the case where a set of parameters at the interface of a component is found to 
lead only to scenarios with higher cost. Then, all the scenarios with identical value assignments for these 
interface parameters could have their A 2 ' cost estimate increased to anticipate the higher cost expected to be 
incurred during further expansion. This is not currently implemented and would be subsumed to a large ex- 
tent by a scenario recombination mechanism described in Section 12 on future work. 

The dynamic adjustment of the cost estimator is an instance of dependency-directed backtracking because 
the cause of the low quality scenarios is looked up higher in the search tree, appropriately modified, and in- 
hibits further exploration of the afflicted branch; at least until no better options are left. The same mechanism 
could be used to eliminate all partial scenarios which share parameter assignments which can be shown to lead 
to inconsistencies. Inconsistencies are detected in step 4 of the algorithm presented below, when no successor 
scenarios can be generated. We are now ready to define the diagnostic search algorithm D-Search. 

Algorithm D-Search: 

1 . Create an initial active scenario S*. All its parameters map to Unknown and all the components map to 
ToBeAnalyzed. Set the set AS of active partial scenarios to {S t }. 

2. Fill in the known data: classify measurements into qualitative values and set the values of the measured 
parameters in 5 accordingly. Make the initial scenario S« the current scenario S* 

3. Choose a component compjhom the current scenario S t which is mapped into ToBeAnalyzed by func- 
tion Mode of Si and apply the expansion operator associated with its component type. Remove S, from 
AS. 

4. If no successor scenarios were generated in step 3, i.e. the set of parameter assignments and behavior 
modes of S,- is inconsistent, then goto 6, else apply the heuristic evaluation function to the successor 
scenarios. 

5. If the successor scenarios {SJ, k = l,...,m, where mis the number of successors generated in step 3, 
are finished, then add them to FS, the set of finished scenarios, FS "= FS u {Sj, else add them to AS. 

6. If AS is empty, then goto 10, else rank the active partial scenarios in AS according to cost /. 



Final Report July 1993 19 



Enhancements to the Engine Data Interpretation System 
(EDIS) 

7. Select the best (lowest cost / ) partial scenarios from AS and make it the current scenario S,. Break ties 
arbitrarily. 

8. Global consistency check: Find the component in S; analyzed last. Check that its interface parameters 
are assigned one value v e Val- Unknown only. If there are conflicting values assigned, then remove 
Si from .AS and goto 6. 

9. Goto 3 

10. If FS is empty, then no consistent scenarios could be found and the algorithm failed to generate a diag- 
nosis; otherwise the scenarios in FS enumerate all possible behaviors and thus all possible faults. 

It should be noted that the set FS is likely to be very large, especially in the situation of interest where few 
parameter values are known. The "best" diagnosis is not necessarily minimal, though. The heuristic evalua- 
tion function can be tuned to prefer a combination of several faults over some single faults. This feature is 
useful when secondary faults may be induced by a primary fault. 

The set FS is empty only in the exceptional case when the measured values are inconsistent with any possible 
device behavior. This case may occur when sensors malfunction or measurements are incorrectly interpreted. 

The last issue to be addressed concerning the diagnostic search algorithm is step 3, the selection of a compo- 
nent to be analyzed next. Component selection determines the order in which the search space is explored. If 
the component which is actually faulty is chosen early, then the algorithm will produce the correct diagnosis 
fast. At this step, additional expertise should and can be brought to bear on the diagnostic search. The diag- 
nostic system which has been developed around the proposed algorithm can execute a set of heuristic rules or 
request user input to select a component to be analyzed and also a behavior mode to assume. When a user 
chooses to submit his or her own hypothesis, the system will test whether it is consistent with the available 
data and evaluate its quality relative to competing hypotheses. 

Components are selected based on the number of unknown parameter values associated with each compo- 
nent, reasoning focus and continuity control, and the likelihood that this component is the cause of the anoma- 
lies. The number of unknown parameters is used to estimate the number of different allowable value 
assignments to the remaining parameters. Fewer unassigned values usually imply stronger restrictions on the 
remaining parameter values and thus a higher likelihood of choosing the correct value. Fox [3] has formalized 
this heuristic as "variable value goodness texture." 

A generic constraint satisfaction algorithm might use a search process with the single goal of optimizing 
search efficiency. In an interactive system the user who monitors reasoning progress has to be considered. 
Users more easily follow depth-first search which fully explores a single line of reasoning than an optimized 
strategy which appears to jump between various lines of reasoning based on different assumptions. 

Once a set of components has been analyzed, the algorithm will tend to select a component for analysis which 
is connected to the component which was analyzed last. The reasoning thus follows the structural intercon- 
nectivity of the device as represented by the device schematic, emulates a human expert reasoning strategy, 
and facilitates explanation of system behavior. 

The most effective way to streamline search and constraint satisfaction is to identify the faulty component as 
early as possible and to guess its fault mode correctly. After that, choices are limited to correct behavior 
modes for all other components (in the common case of a single fault), of which there exist far fewer than 



20 Final Report, July 1993 



Enhancements to the Engine Data Interpretation System 

(EDIS) 

faulty behavior modes. Two somewhat different operating modes of the constraint satisfaction search can 
thus be discerned. An 'exploring* mode, which is in effect before any fault assumption has been made and the 
search tries to locate the component which may account for the anomalies (V -= 5), and a Verification' 
mode, which is entered after a component has been incriminated and the constraint satisfaction algorithm tries 
to show that the assumption is consistent with the data and does not require unlikely assumptions about other 
component behaviors (A/ = l). 

Guessing the responsible fault behavior, i.e. hypothesizing which component is faulty and is causing the ob- 
served anomalies, is supported by any and all of heuristic expertise, component failure rates, and probabil- 
ities of specific faults, if available. Heuristic functions are not restricted to investigate only data associated 
with the component under considerations, but may take overall system behavior into account. In a feedback 
system, such as the SSME, telltale effects of faults can sometimes be observed at sensors far removed from the 
original cause. 

The algorithm presented above concentrates on a single hypothesis at a time and implements a single line of 
reasoning. Multiple hypotheses could be explored in parallel either by choosing more than one "best" partial 
scenario in step 7 or by applying more than one operator in step 3. The algorithm presented appears to be well 
suited for distributed implementation because such multiple lines of reasoning require very little interaction. 
Information to be shared only travels up and down the search tree. Only the selection of the best partial scenar- 
io is a global operation. Information traveling down the search tree implements the normal line of reasoning. 
Information travels up the search tree when special conditions, such as inconsistencies, are encountered dur- 
ing expansion which are to be considered in the heuristic evaluation functions of scenarios waiting to be ex- 
panded. 



I 



Final Report July 1993 21 



Enhancements to the Engine Data Interpretation System 
(EDIS) 



5- SSME Configuration 

Figure 3 shows a top level view of the SSME configuration. Only the "Terminal Components" are shown 
(defined in file terminal) which do not take part in the reasoning process. They provide the linkage to the 
environment of the SSME. The two pipes F190 and 190 lead to parts of the system which are not modeled. 
Also not modeled are the controllers for the valves MFV, CCV, MOV, and OPOV Only the FPOV controller 
is modeled. The following figures show additional detail of the modeL All components are shown in the form 



Cw&£ 



Object 



where Object is the name of the component and Class is its class name. Some of the names are standard, others 





iiliiiiiliwMiii 




HHUMI 


?MM&8$\ 






FUELTANK 


LOXJTANK 


1 


















- 


;". f^^^^W^:^ 




b c 

a 

h SSME 

g 


d 
e 
f 




:|;; ^tSffiVW$M&f*% £&i3.;li 


F190 






O190 








.-.. :;= -■-... 


^sMM^^^mm 






i^m^^^hm 


MFV CTRL 






OPOV_CTRL 








--- ■■:-■■ 






liPiiKfflii^M 






MOV_CTRL 




















■mHBBI 








CCV_CTRL 





Figure 3: SSME Model - Components and Interconnections, Level 1 



were invented for this project (they can easily be changed later). Figure 4 shows the four main blocks of the 
SSME configuration: fuel and oxygen supplies, the cooling piping, and the MCC and nozzle assembly. The 
MCC/NOZZLE block is shown in more detail in Figures 5, the COOLING block in Figure 6, the FUEL 
SUPPLY block in Figure 7, and the LOX SUPPLY block in Figure 8. Appendix A.2 contains the set of config- 
uration files which define the SSME configuration. 



22 



Final Report, July 1993 



Enhancements to the Engine Data Interpretation System 

(EDIS) 



g | (from CCV_CTRL) 



t-_T 


a 


b (from 

If „, . 


FUELJTANK) 






ol 




(fromLOXJTANK) c 




Ffc^ 


FUEL 
SUPPLY 








LOX 
SUPPLY 


d 


W- 


ml 


MCC/NOZZLE 


m2 




(to 


m 


(to 
F190) 

h 


O190) 


m 




m3 


e 


W 




(from 

OPOV 

CTRL 


%u 


(from 
MFV 
CTRL 









tl 


i 


t2 






= 


c3 


COOLING 


c4 


f 


w? 


c2 


(from 






m 


cl 


MOV 
CTRL 


^ZtT^ 























Figure 4: SSME Model - Components and Interconnections: Block Dia- 
gram 



ml 



(from HPFT) 



A i 



HGM 



B 



K6 ,:)j(Hp 



'05 :.wjair* pMwI^^IP 



MCC 



(to MCC_COOL) 






NOZZLEl 



tl 



m2 



(from HPOT) 



m3 



(from MOV) 



Note: exhaust into atmosphere 
not modeled 



t2 (to NZLCOOL) 



Figure 5: SSME Model - MCC/NOZZLE 



Final Report July 1993 



23 



Enhancements to the Engine Data Interpretation System 
(EDIS) 



tl (from MCC) 
(to F109) 


i 


t2(fromNOZZLEl) 






cz 


il^^Ml^^^^M 


/ 


.• r pcpj^^silil 


B 


iiiiiMii^iiiiiii 






MCC_COOL 


NZLOOOL 


* 


MIXER 






J 


C ] 


/ 




A 


L 






liiiiipiiii^iiiii 


/ 


: $$&%&. ^££^mQS £i 


/ 






cl 


DIFFUSER 


A 


ccv 


F107 




(fromMF 


V) 






u 




J 






B> 


iliKiWiiiiiPi 




g (from CCV_CTRL) 


M103 










/ 


A < 


r . 






■ :*|m:j: : .%%^:;^^ 


• 


illiiii^Kiiiiiiii 






F110 


F108 




(to 
Figure 6: SSME I 


FPB) c3 

4odel - COOLING 




c4 (toOPB) 



c2(fromMCC_COOL) 



M-il-.iiM'iJira.-iijIi'w 



F109 



(to F190) b 
a 






LPFT 



wmw-KtiWifflmffl-mafiw 



MYt'iVi'iViYiYiviV 



X'lvKa-fflgjffgf^g^gJigJ:' 



LPFP 



(from FUELJTANK) 



F101 



(fromO206) 




c3 (fromFHO) 



HPFT 



ml (toHGM) 



ffiffiffigTOMM 



HPFP 



Figure 7: SSME Model - FUEL SUPPLY 






(to DIFFUSER) 



MFV 



Cl 



h (from MFV_CTRL) 



24 



Final Report, July 1993 



Enhancements to the Engine Data Interpretation System 

(EDIS) 



m 



(toFPOV) 
ol 



Mmttixmm!& 






O206 



(from OPOV_CTRL) 
c 







OPOV 



'iM^^^^^^^MmM^M 



OPB 



m2 (toHGM) 






^m 



HPOT 



(fromMOVCTRL) 
f 



rn3 
(to MCC) 






MOV 



mm 



O204 





| (fromLO XJTANK) 
c 



:«ow^««;K«<gM5S 



:.m^»/A.a/g^.yg/.flvA.-: 



LPOP 






I 



O201 






■".:."->".:o:.>-o:^>-.:-.->>!--.:...: 



0203 



Figure 8: SSME Model - LOX SUPPLY 



Final Report July 1993 



25 



Enhancements to the Engine Data Interpretation System 
(EDIS) 

6* Heuristic Rules 

Traditional expert systems rely heavily briUie ability of heuristic ruleslo rapidly identify rommon faults. 
Finding common faults fast is also a design goal of EDIS but, in addition, broad fault coverage is desired. 
Common faults can be dealt with by providing heuristic rules which, when successfully executed, predict a 
likely fault and force EDIS to prefer those assumptions about SSME behavior which are consistent with the 
fault predicted by the heuristic rules. This feature must be explicitly enabled by setting the slot USEJHEU- 
RISTIC_RULES.Value to TRUE. Unlike the rest of EDIS these rules are of necessity specific to the SSME 
and are therefore stored in a file in the confssme directory with the other SSME specific configuration data. 
There are also two more files (fuel_slde and Iox_side) which describe a grouping of components specific to 
the SSME. These files are read only when^lot USE_HEURISTIC_RULES. Value is TRUE. 

A small number of heuristic rules are implemented which can identify certain likely faults rapidly. These 
rules may be run at the start of the reasoning process and fundamentally change the way the search space is 
explored. Without heuristic suggestions, EDIS tries to determine system behavior and hypothesizes faults 
only to satisfy specific anomalous parameter values. If a heuristic rule identifies a likely fault, EDIS attempts 
to find a behavior which is consistent with this fault assumption. Heuristic rules easily combine evidence 
from different parts of the system. For example, the heuristic rule whicSidehtifies a leak in the MCC Cooling 
duct tests for anomalous values at the LPFP as well as the MCC Cooling duct. In some other cases rules check 
for effects of fuel side problems at components as remote as the HPOTP. On the other hand, behavior synthe- 
sis proceeds component by component and uses only data local to a component to generate fault hypotheses. 
Given a set of well designed heuristic rules, a hypothesis generated by a heuristic rule will most likely be 
correct, while hypotheses generated based on local behavior only are more tentative and are likely to be found 
to be inconsistent with the remaining data. 



Unfortunately, it is still not obvious which one of these two methods will arrive at an answer earlier. In the 
example included, the standard diagnostic process which does not utilize heuristic rules finds the correct an- 
swer earlier even though it selects and discards several wrong hypotheses before it generates the correct one. 
The problem with the heuristic suggestions is that they force EDIS to develop a consistent behavior starting at 
a component with many unknown parameters. Many unknown parameters will lead to many possible behav- 
iors because the leak hypothesis is not specific enough to effectively limit the number of possible behaviors. 
The standard reasoning process chooses components for investigation in an order which minimizes guessing 
and is therefore likely to derive the correct behavior which implicitly contains and ultimately reveals the cor- 
rect fault hypothesis. The reasoning process which executes heuristic rules has not been refined as much as 
the standard constraint satisfaction approach. Several modifications are possible which could improve the 
performance of EDIS when given a heuristic suggestion which could not be implemented yet. Fortunately, 
most of them also promise to enhance the standard reasoning process. 

The shortcomings associated with using heuristic rules can, at least sometimes, be overcome by using data 
generated by the Power Balance Model (PBM), see below. This is not, however, a guaranteed way of effi- 
ciently solving the diagnostic problem, but just another heuristic method which, we hope, will work most of 
the time. 

Below a listing of the implemented heuristic rules can be found, formulated in structured English. File 
heuristic-rules . tkb included in Appendix A.3 contains a listing of the rule code in NEXPERT syn- 
tax. EDIS will use the heuristic rules only if the slot USEJHEURISTIC_RULES.Value is set to TRUE be- 



26 Final Report, July 1993 



mi 



Enhancements to the Engine Data Interpretation System 

(EDIS) 



fore processing starts. Rules may either suggest a specific fault in a specific component, such as LOW 
EFFICIENCY of the HPFP, or a specific fault in a class of components, such as LEAK in a FUEL-SIDE 
DUCT, i.e. a pipe or cooling duct. 

if 



HPOT discharge temp 


is 


HIGH 


and HPFT discharge temp 


is 


NORMAL or HIGH 


and HPFP discharge pressure 


is 


LOW 


and HPFP speed 


is not 


LOW 


and MCC pressure 


is 


NORMAL 


then suspect a 






LEAK 


in a 


FUEL-SIDE DUCT 



Rulel 



if 



HPOT discharge temp is 

and HPFP discharge pressure is 

then suspect a 

LEAK in a 



HIGH 
LOW 

FUEL-SIDE DUCT 



Rule 2 



if 



FPOV position 
and HPFT discharge temp 

and FPB pressure 

then suspect a 

LOW EFFICIENCY 



is HIGH 

is HIGH 

is HIGH 

in the HPFP 



Rule 3 



if 



LOW 
LOW 
LOW 



LPFP speed is 

and MCC Cooling disch temp is 

and MCC Cooling disch press is 

then suspect a 

LEAK in the MCC Cooling duct 
Rule 4 



if 



and 



HPOT discharge temp is HIGH 

MCC Cooling disch temp is HIGH 



Final Report July 1993 



27 



Enhancements to the Engine Data Interpretation System 
(EDIS) 



and LPFP speed 

then suspect a 

LEAK 



Rale 5 



is 



LOW 



in the NOZZLE Cooling duct 



If 



OPOV position 


is 


HIGH 


and HPOTP speed 


is 


NORMAL 


and MCC pressure 


is 


NORMAL 


then suspect a 






LOW EFFICIENCY 


in the 


HPOT 


Rule 6 






ific component is implicated, the search 


process starts at this component, finds all po 



haviors of the component and identifies those component behaviors which are consistent with the fault hy- 
pothesis. One of these behaviors is chosen and EDIS tries to find a consistent behavior of the whole SSME 
given the fault assumption and the assumptions made when developing the initial component behavior. 
When a set of components is implicated, EDIS initiates the search in standard order, i.e. with components 
where few assumptions have to be made. When a component is encountered during search which is in the 
implicated set and one or more of its behaviors are consistent with the hypothesized fault, the fault is assumed 
to have occurred at thFs component. EDIS then ^continues in its attempt to justify thelault hypothesis by com- 
pleting the behavior of the SSME. When a set of components are implicated, they are thus tried in the order in 
which they are encountered during the search, at least until one of them forms the basis of a complete and 
consistent SSME behavior. 



7, Power Balance Model 

The Power Balance Model performs data reduction after engine tests. A file is produced which contains val- 
ues for many internal immeasurable parameters. Some of these parameters are used within the EDIS qualita- 
tive model, too. A method was developed through which the results of the PBM-based analysis can guide the 
heuristic search performed by EDIS. This feature must be explicitly enabled by setting the slot 
USE_PBM_DATA.Value to TRUE. The file PBM_values.nxp must be created before enabling this feature. 
It contains definitions of "template" objects whose parameter values have been filled with the available PBM 
data. EDIS compares the component behaviors it generates against these template objects and rewards those 
behaviors which have larger numbers of matching parameter values. 

Parameter values generated by the PBM do not convey the same level of confidence as measured parameters 
because the PBM has only limited fault simulation capabilities. EDIS therefore does not add the PBM sup- 
plied parameters to the set of measured parameters but only uses them to identify likely SSME behavior. Any 
component behavior generated by EDIS during the search is compared against the PBM predicted values and 
the better the match the greater the chance that the proposed behavior represents the actual behavior. The 
current implementation subtracts 0.02 from the local quality of any behavior for each parameter value which 
does not agree with the PBM data. 



28 



Final Report, July 1993 



I 4 Enhancements to the Engine Data Interpretation System 

^ (EDIS) 

— The PBM generates numerical data which are translated into qualitative values via a process described below. 
Since we did not have any better information we set the limit above which a deviation would be considered 

~ anomalous at 2.5%. Performance of EDIS is quite sensitive to this limit. For example, we selected only one 

~ of several data for MCC cooling duct flow. One of the values showed a 2.42% increase, another a 2.56% 

^ increase (at locations 1103 and 1104 in the A-ARRAY, respectively). Using the larger of the two, EDIS per- 

^ formed as expected since larger than normal flow is consistent with a leak. Using the lower deviation value, 

** EDIS was unable to verify the leak hypothesis in reasonable time. This "brittleness" of performance is typi- 

cal of "crisp" qualitative classification (and also of traditional heuristic rule-based systems). Preliminary 
™ results on our research into the application of fuzzy classification to SSME diagnosis are described in Section 

8. Fuzzy classification promises to alleviate the brittleness problem. 

5 The use of PBM data does not significantly change the search process in our example. The PBM predicts 

^ most parameter values to be normal and EDIS already favors normal values over deviations for normal com- 

= ponent behaviors. The search therefore proceeds exactly as it does without the use of PBM predictions. There 

S are small differences in some of the behaviors but these do not alter EDIS * interpretation of component behav- 

ior modes. 

5 A combination of heuristic rules and PBM data proved to be the most effective way of diagnosing the MCC 

cooling leak fault in the example. The heuristic rules correctly predict the MCC cooling leak and the PBM 

_ data correctly predict values for three critical parameters. These predictions combined with some measured 

H values lead EDIS to select the "correct" behavior for the MCC cooling duct and the leak is diagnosed in a 

single pass through the component network without any backtracking. It is interesting to note that some de- 

= tails, i.e. parameter values, differ in the answers generated by EDIS in its normal qualitative search mode and 

iJ the heuristic/PBM guided mode. Roughly speaking, the heuristic/PBM solution corresponds to a bigger leak, 

e.g. the outflow is assumed to be LOW, while the qualitative search predicts a small leak, e.g. the outflow is 

= assumed to be NORMAL. Both solutions comply with the low value of output pressure measured by a sensor. 

m Either assumption is consistent with lower than normal mechanical power generated by the LPFT. 

— The directory edis3/PBM contains executables and sample files which illustrate (he creation of the PBM_va- 
V lues.nxp file. The file PBM_parameters must be available. It contains a subset of the entries of the 

PBM90A A- ARRAY variable listing from file vardoc (from EPVAX) as of 18-Dec-1992. Lines with pa- 
= rameters which have equivalents in the EDIS qualitative model contain a bracketed term at the end of the line 

BJ which indicates the corresponding component and the component parameter. For example, the line 

s 4 P1FP1 LPFP INLET PRESSURE [LPFPpin] 

indicates that PBM parameter P1FP1 at location 4 is equivalent to the input pressure (p in ) parameter of the 

— LPFP in the qualitative model. The list of equivalences is as yet possibly incomplete. 



Next, the files containing the test data and the comparison data must be moved into this directory. They are 
currently stored on the IBM system. The comparison data file may be off-line and has to be loaded. Now 
program displaydb can now be executed to prepare the intermediate file PBM_numeric_deviations. 

Execute program displaydb using the commandline 

displaydb comparison jlatajile testjiatajile timejlice -1 > PBMnumericdeviations 

to create file PBM_numeric_deviations. The sample file was created with 



Final Report July 1993 29 



Enhancements to the Engine Data Interpretation System 
(EDIS) 



displaydb al613 al614 29 -1 > PBM_numeiic_deviations 

Individual values can be displayed using an alternate form of this command. 

displaydb testjiatajile timejslice index 

Using time slice 1, the test date and test duration and the time slice duration can be displayed at indices 934, 
935, and 936, respectively. After selecting a time slice, displaying the datum at index 937 reveals the start 
time of the chosen time slice. 

Finally, program make_PBM _params can be executed. It reads files PBM_parameters and PBM_nu- 
mericdeviations and writes file PBM_values. nxp. This file must be moved into the edis3/confssme direc- 
tory. No parameters are necessary since all files have standard names. All files for the sample case described 
above are listed in the appendix. 

8. Fuzzy qualitative system 

A mathematical model of a system describes the system in terms of the underlying analytical equations that 
determine its behavior. It is required to know the exact relations between system variables to develop a math- 
ematical model. A mathematical model is an exact representation of the system and produces exact results. 
Analytical equations do not represent knowledge about the system explicitly. Commonsense knowledge a 
person has about the system, cannot be represented in a mathematical model. A mathematical model suffers 
from the "interpretation problem." 

A qualitative model provides an alternative to a mathematical model in a complex and uncertain environment. 
In the absence of exact analytical equations, an abstract qualitative model can be developed. Given limited 
numerical information about the system, a qualitative model can produce very useful results. If the problems 
are difficult to solve numerically and the precision of the results required is not high, it is advisable to resort to 
qualitative methods. Another advantage of qualitative model, apart from relaxing the requirement of precise 
numerical information, is the ability to represent the commonsense knowledge explicitly and therefore, 
yields easier interpretation. But the intentional neglect of the available numerical information may result in 
over abstraction of the system. The results obtained from an over abstracted model are imprecise. Qualitative 
systems also suffer from the limitations associated with the inherent ambiguity in qualitative arithmetic. Re- 
solving the ambiguities in qualitative arithmetic increases the precision of results and decreases search com- 
plexity. One approach is to explore an unambiguous mathematical formalism for qualitative variables and 
another is to make use of the available quantitative information to refine the results obtained from qualitative 
analysis. 

Fuzzy qualitative modeling is a combined approach which makes use of all the available quantitative in- 
formation and is supported by the arithmetic of possibility theory of fuzzy sets. The Fuzzy qualitative model- 
ing paradigm integrates possibility theory of fuzzy sets with qualitative interval calculus for more detailed 
and accurate modelling rftl^syst^7Tlus reduces the ambiguities inherent in the pure qualitative methods 
and produces more precise results than those obtained by a pure qualitative model. This is a generalization of 
qualitative modeling and offers an intermediate level of model abstraction. 



30 Final Report, July 1993 



ij Enhancements to the Engine Data Interpretation System 

(EDIS) 

" The fuzzy qualitative models can be categorized as deep causal models which capture underlying causal phe- 

nomena and facilitate reasoning from first principles. These models can be used as generic components of a 
pi model-based diagnostic system. 

8.1 Fuzzy qualitative model 

A component model describes all the possible behaviors of the component In a constraint-based model, 
behavior of a component is described by a set of constraints. In a fuzzy qualitative model, the modeling primi- 
-—^ tives are fuzzy constraints and fuzzy qualitative states. 

8.1.1 Fuzzy qualitative states 

|g Possibility measure is a natural way of representing subjective uncertainty. It is the measure of material diffi- 

M culty of an event occurring plus the subjective evaluation of the occurrence of the event [8]. To model the 

uncertain belief, it is required not to rigidify the relationship between the indications one has in favor of an 
if event and those that weigh against it. Unlike probability, possibility of an event is independent of the possibil- 

^ ity of the contrary event. 

|j| The range of a fuzzy variable is a closed interval bounded by the maximum and minimum possible values, 

W chosen to ensure that the whole range of interesting behaviors is covered. The closed range of the variable is 

divided into an arbitrary but finite number of fuzzy subsets or fuzzy intervals. Each fuzzy interval represents a 
— fuzzy qualitative value. The set of fuzzy qualitative values covering the whole range of interest allows all 

«* numerical values that the variables may take to be mapped onto their associated fuzzy qualitative values. The 

number of fuzzy intervals chosen depends upon the granularity desired. Since the range is divided into a 
finite number of fuzzy intervals, a variable takes a finite number of fuzzy qualitative values. There is a direct 
mapping from numerical range to fuzzy qualitative values. A variable has, associated with it a quantity space 
M Q, with the following properties. 

H * finiteness: The range is divided into a finite number of fuzzy intervals and therefore a variable can take on a 

finite number of fuzzy qualitative values. 

U • Coverage: Fuzzy qualitative values the variable can take on cover all the behaviors of interest. 

• Mapping: There is a direct mapping between the numerical range and the fuzzy qualitative values. 

^2 • Granularity : The number of fuzzy intervals is arbitrarily chosen depending upon the granularity desired. 

" • Closed: The range is closed and all the possible numerical values outside the range can be conveniently 

== represented by the fuzzy qualitative values of the intervals at both the ends of the range. 

If • Overlapping: Fuzzy intervals are overlapping to account for ambiguity in the definition of fuzzy qualita- 

tive values. 



U 



A fuzzy interval, for example the value"Low", is defined by a fuzzy number which is represented by a 4-tu- 
ple (a, b, c, d), where p(a) = 0, p(b) = 1.0, p(c) = 1.0, p(d) = 0. p(x) is the possibility of a numerical value x 
falling in the qualitative value ("Low" in this case). The 4-tuple representation implies a trapezoidal shape 
for the possibility distributions of fuzzy numbers, in general, any convex function could be chosen. 



Final Report July 1993 31 



Enhancements to the Engine Data Interpretation System 
(EDIS) 




a b c d 

Figure 9: A fuzzy interval 

The fuzzy qualitative domain consists of n values if there are n fuzzy intervals in the range. The value of a 
fuzzy variable is represented by an n-tuple(Pi, P2, ... P n ) where Pj is the measure of possibility in the ith inter- 
val. 

For simplicity, consider a fuzzy qualitative domain with only three fuzzy intervals, low, normal and high. 
Heuristic information is required to define the fuzzy intervals. In the present example, the "normal" fuzzy 
interval is defined symmetric withfespect to the origin. The intervtfs, low and high are defined symmetrical 
with respect to the normal interval. It is common to select overlapping fuzzy intervals in such a way that the 
sum of the possibilities of a numerical value falling in any of the fuzzy intervals is always equal to 1. 

This mode of representation of the value of a variable enables a convenient mode of switching between the 
fuzzy qualitative method and the crisp qualitative method. The subjective definition of the fuzzy sets 
introduces an element of subjectivity. 



Low 



Normal 




High 




Figure 10: Definition of three overlapping fuzzy intervals 

8.1*2 Fuzzy interval arithmetic 

Fuzzy interval arithmetic is a generalization of interval arithmetic. Fuzzy interval arithmetic operates on the 

from the possibility values of the fuzzy inter- 



possibility values of fie interval and not the intervals direc 
vals, a corresponding interval on the real number line is computed and the arithmetic is done over these inter- 
vals. The possibility values of the resultant interval in all the fuzzy intervals of the domain are computed. 



32 



Final Report, July 1993 



S 



Enhancements to the Engine Data Interpretation System 

(EDIS) 



Fuzzy qualitative addition: Consider a simple example of adding two variables with fuzzy qualitative values. 
Let A, B and C be the variables, whose fuzzy values are defined by the fuzzy intervals as shown in Figure 11 
below. 



A(ffigh) = (3, 6, 8, 8) 
A(Noimal) = (- 6, - 3, 3, 6) 





A(Low) =(-8,-8, -6, - 3) 



8-6 -3 36 8 

Figure 11a: Fuzzy intervals for variable A 




B(High) = (2, 4, 6, 6) 



B(NormaI) = (-4,-2,2,4) 



B(Low) = (- 6, - 6, - 4, - 2) 



-6 -4 - 2 



Figure lib: Fuzzy intervals for variable B 





C(H) = (5,10,14,14) 



C(JV) = (- 10, -5,5,10) 



- 14 - 10 - 5 5 

Figure lie: Fuzzy intervals for variable C 



10 14 

C(I) = (- 14, - 14, - 10, -5) 



To find the sum of A(High) and B(Low): C(?) = A(High) + B(Low) 

A(High) =(0.0,0.0, 1.0)... possibility of falling into interval Low (=0.0), Normal"(=0.6), and High (=1.0) 
=(3, 6, 8, 8)... fuzzy number representing fuzzy interval High of variable A; 



Final Report July 1993 



33 



Enhancements to the Engine Data Interpretation System 
(EDIS) 

B(Low) = (1.0, 0.0, 0.0) = (-6, -6, -4, -2); 
A(High) + B(Low) = (-3, 0, 4, 6) 

The resulting sum, which is again represented as a fuzzy number, is now mapped in the quantity space of 
variable C. Find the area overlapped by the resultant interval onto the three fuzzy sets of C. The ratio of the 
area overlapped on each fuzzy set to the total area of the fuzzy set corresponds to the possibility of the result 
lying in that set. The above resultant interval gives the following possibility figures: 
C(0.0, 0,43, 0.01) « C(0.0, 4.3, 0.0). 

Unlike qualitative addition, fuzzy qualitative addition is well defined and does not give ambiguous results. If 
the result has non-zero possibility values in more than one fuzzy set, either only one fuzzy set with maximum 
possibility value can be considered or all the fuzzy sets with possibility greater than a preset limit. The sym- 
metric definition of the fuzzy sets guarantees the existence of the additive inverse. 

Fuzzy constraints are bi-directional. To check for the consistency of the solution in the previous example, 
find the value of A, given C and B. Use the fuzzy sets and the possibility values of the previous example. 
A(?,?,?) = qo.0, 0.43, 0.0) - B(1 .0, 0.0, 0.0) 

= (-10, -7.15, 7.15, 10) - (-6, -6, -4, -2) 

=(-4, -1.15, 11.15, 12) = (-4, -1.15, 8, 8) 
This results in A(0.0, 0.81, 1.0). Note that A(0.0, 0.0, 1.0) is one of the solutions. It can be seen from the 
non-zero value of the possibility for the value of A falling into fuzzy interval "Normal," that fuzzy calculus 
introduces some ambiguity, but not as much as crisp qualitative calculus. 

8.1.3 Fuzzy Constraints 

Fuzzy constraints are abstractions of the algebraic constraints that determine component behavior. Fuzzy 
constraints are relations between fuzzy qualitative variables. The factor of satisfaction of a fuzzy constraint 
can be graded. Testing for a fundamental fuzzy constraint is essentially comparing two fuzzy numbers corre- 
sponding to the left hand side and the right hand side of the constraint. 



Measuring equality between two fuzzy numbers: 

The difference between two fuzzy sets can be found by summing the squared differences between them. 
Normalizing this result by dividing by the support value ( the maximum support minus the minimum sup- 
port) results in a grade. This grade reflects how different the two fuzzy sets are. Negating this grade results in 
a grade for how equal the sets are. If the result is 0.0 then the sets share no members (to any degree). If the 
grade is 1.0 the two sets are identical. This grade of equality can be taken as degree of satisfaction of the fuzzy 
constraint. 

The grade of equality between two fuzzy numbers A(ai, a 2 , a3, ^4) and B(bi, b 2 , b3, b4) \ 

= 0>i -ai) 2 + (b2 -a 2 ) 2 + (b 3 -a 3 ) 2 + (b 4 -a 4 ) 2 7 ( max_diff - min_diff) 
where 

max_diff = max ( (bi -ai> (b 2 -a 2 ), (b3-a 3 )' (b 4 -a4) ) 

min_diff = min ( (bi -a^* (b 2 -a 2 ), (b 3 -a 3 ) > (b 4 -a4) ) 

Quality of a component beh avior: r ^^^ ; ^ r r ^ _ , r _ 

The quality of the behaviors can be computed using the degree of satisfaction of the constraints. A component 
behavior is normal if the degree of satisfaction of all the fundamental constraints is 1.0. The quality helps in 
ranking the behaviors and selecting the behavior with best quality for further expansion. 



34 



Final Report, July 1993 



Enhancements to the Engine Data Interpretation System 

(EDIS) 



8.2 Implementation 

The fuzzy qualitative diagnostic system for fault diagnosis of SSME is being implemented in C++ in the 
HP-UX environment. An object diagram of the diagnostic system for SSME is shown in Figure 12 below. 




Figure 12a: Object Diagram 

The fuzzy qualitative space for all the parameters is implemented as object class "Fuzzy_Value." The class 
Fuzzy_Value has the following features: 

• it defines the three fuzzy sets, low, normal and high using three fuzzy numbers, 

• it stores the fuzzy values which represents the possibility values in the three fuzzy sets, 

• it stores the numerical range of the variable, a 4-tuple number, 

• it defines the mathematical operations addition, subtraction, multiplication, and average, 

• it defines an equality relation for objects of class Fuzzy_VaIue. 

All the parameters in the SSME are abstracted under a common class, Parameter. A parameter has an 
associated name if it is measured. A parameter maybe either measured or derived. The value of a parameter is 
compared against a comparison value and the deviation is mapped to its fuzzy qualitative space. The values of 
interface parameters are common to the neighboring components and are propagated between the neighbors. 



Final Report July 1993 



35 



Enhancements to the Engine Data Interpretation System 
(EDIS) 



Class Port: 

A thermodynamic component has one or more ports. A port has three parameters, pressure, flow rate and 

temperature. A port is shared between two neighboring components. 

Component interconnections are implemented as objects of class "Interconnections" which consists of the 
names to the two neighboring components, A and B, and their corresponding ports, port_A and port_B. The 
set_port_A function propagates interface parameters from the port_B of component B. The set_port_B func- 
tion propagates interface parameters from the port_A of component A. 

A component is an aggregate of the constituent ports, interconnections, derived parameters and the corre- 
sponding list of behaviors created. Each component xxx has a class xxx_behavior defined for it. 



o 



Component 



Port 



^y 



^j 



o 



Interconnection 



Component_behavior 



Parameter 



Figure 12b: Component object class 

Class "Behavior" is an abstract class for all the component behaviors so that they can be grouped together in 
one collection. A component_behavior has a mode and a list of assumptions. The mode of the component 
behavior may be either normal, or any of the fault modes. A mode is characterized by quality which is a mea- 
sure of the satisfaction of the fundamental constraints. 

Class "Scenario": A scenario is a collection of the behaviors of the analyzed components. It stores the name 
of the last analyzed component. Class Scenario has a cost attribute. Each scenario is ranked based on the cost 
function. A scenario is chosen for expansion when it has the lowest cost function value. 

Class "Example" is an abstract class that acts as superclass to the subclasses - Scenario, Behavior, Compo- 
nent, Interconnection and parameter classes. It consists of all the methods of Scenario, Behavior, Component, 
Interconnection, Parameter classes. 

The functional diagram of the diagnostic system for the SSME is shown below: 

1. Scenario consists of all components, which are yet to be analyzed. Read in the configuration files, one for 
each component to initialize the structure of SSME. Read in the fault modes of each thermodynamic compo- 
nent. 

2. Read in the measured data and the comparison data. Find the possibility value of the parameters in all the 
three fuzzy sets. 

3. Choose a component compj from all the components with maximum ratio of number of known parameters 
to unknown parameters to analyze its behavior. 



36 



Final Report, July 1993 



Enhancements to the Engine Data Interpretation System 

(EDIS) 



^ 



Read in the structure of the SSME - components 
interconnections, fault modes 



T 



Read the measured dataand comparison data 



10 



Find the scenario 
to be expanded 




Choose a component for analyzing 



Contradition found: 
Find the inconsistent val- 
ues of interface parame- 
ters and assign high cost 
to the scenario. 
Find the assumptions 
which caused the assign- 
ment of these values. 



Find the expected mode of the behvaior 
of the component to be analyzed 



Pick two most possible behaviors of the chosen 
mode for the component to be analyzed 
Set Component = Analyzed 
Find the quality of behaviors 



Propagate the values of 
interface parameters of 
the analyzed component 



Figure 13: Data flow diagram 





e 



Final Report July 1993 



37 



Enhancements to the Engine Data Interpretation System 
(EDJS) 

4. Find the expected mode of the behavior for the component chosen for analyzing. 

5. Pick the two most possible behaviors of the chosen mode for the component to be analyzed. Set the ana- 
lyzed flag of the component. Find the degree of satisfaction of the fundamental constraints and compute the 
quality of behavior 

6. Find the heuristic evaluation function of successor scenarios. If the successor scenarios finished, set the 
finished flag, else add them to the set of active scenarios. 

7. If no behavior was generated in the step 5, Go to step 8. Otherwise go to step 11. 

8. If behaviors of all modes of the component have been tried, there exists no physically possible behavior for 
the component. There exists a contradiction. These particular set of parameter values do not define any com- 
ponent behavior. Go to step 9. If behaviors of all modes of the component have not been tried, go to step 4 to 
pick up the next expected behavioral mode. 

9. Contradiction is found. Find the inconsistent values of parameters which resulted in contradiction and as- 
sign high cost to the scenarios with this particular assignment of values to parameters. Find the assumptions 
which caused the assignment of these values and mark them bad. 

10. Rank the active partial scenarios in AS according to heuristic evaluation function value. Select the partial 
scenario with least cost for expansion from the set of active scenarios, 

11. Global consistency check: The interface parameters of the last analyzed component are propagated to the 
neighboring components. 

12. If all the components are analyzed, stop. Else go to step 10. 

8.3 Comparison against crisp qualitative method 

A fuzzy diagnostic system offers better solutions compared to the crisp qualitative system. 

1. In a pure qualitative system, the deviation of the parameter values can be either low, normal and high which 
indicate only the sign of the deviation of the parameter value and not the magnitude of the deviation. Any 
numerical information, such as the upper and lower bounds of the deviation of a measured parameter, are not 
made use of and therefore it is vaguely represented. In a fuzzy qualitative system, the deviation of parameter 
value is represented by the possibility values in the three fuzzy intervals. The range between maximum posi- 
tive and maximum negative deviations of the parameter value is divided into three fuzzy intervals, low, nor- 
mal and high. This mode of representation takes both the sign and magnitude of the deviation into 
consideration. 

2. In a pure qualitative system, only the qualitative values are propagated between the interface parameters. 
The numerical information associated with the parameter deviation value is neither explicitly represented nor 
propagated. The unmeasured parameters can only have qualitative values with no corresponding numerical 
values and the degree of possibility is restricted to only binary values. 

In a fuzzy qualitative system, the numerical range corresponding to the possibility value is also propagated. 
This results in estimating the partial numerical ranges of the unmeasured parameters, when their values are 
propagated from the neighboring component through interface parameter. 



38 Final Report, July 1993 



Enhancements to the Engine Data Interpretation System 

(EDIS) 

3. Qualitative calculus is inherently ambiguous and lacks additive inverses and multiplicative inverse. The 
direction of solving the qualitative constraints is fixed. Given the values of constraining variables, it is pos- 
sible to find the value of the constrained variable. In general, it is not possible to find the unknown value of the 
constraining variable, given the value of constrained variable. 

In the fuzzy qualitative calculus, by defining the fuzzy sets to be symmetrical with respect to the normal set 
about the origin, it is possible to find the additive inverse of a fuzzy interval. The closed intervals enable solv- 
ing the constraints in both the directions and the consistency of the sets of possible solutions can be proved. 

4. In a pure qualitative system, there is no a>ncept "of partial fulfillment of the constraint. As a result, a large 
number of behaviors may belong to a single behavioral mode. Using fuzzy constraints it is possible to find 
the degree of satisfaction of the fundamental fuzzy constraints which can be used to compute the quality of the 
behavior. Behaviors are ranked using quality factor. IJsing two-place fundamental constraints, finding the 
degree of satisfaction is equivalent to measuring the grade of equality of two fuzzy numbers. Quality of a 
behavior is equal to 1 if all the fundamental constraints are completely satisfied. 

8.4 Limitations of fuzzy qualitative model 

Fuzzy qualitative system requires that more numerical information be given, like the absolute value ranges, 
the maximum and minimum deviations of the parameter values and all the numerical information required to 
solve algebraic equations. Li the absence of the numerical information, it reduces to a simple qualitative sys- 
tem. 

Heuristic knowledge is required to properly define the fuzzy sets, which guarantees that the additive inverse 
of a fuzzy interval can be found. 

8.5 Management of complexity by selective expansion 

In the current implementation, the search space is exhausted completely i.e. all the successor scenarios of a 
scenario are generated. All the possible behaviors of component are created. In a fuzzy qualitative diagnostic 
system, either only the most possible behavior of the chosen mode or a set of behaviors of the chosen mode 
with quality greater than a preset limit can be created. Mode of the behavior to be created is chosen depending 
upon the heuristic information or the global quality of the scenario. A selective expansion of the search space 
is done rather than an exhaustive one. This selective expansion avoids the search space getting unmanageably 
large. 

9. Running EDIS 

EDIS requires a set of configuration and support files located in a configuration directory. An example of 
configuration files which define the current model of the SSME can be found in Appendix A.2. If heuristic 
rules and PBM data are to be used, the corresponding knowledge base and data files must be placed in the 
configuration directory, too. The sequence of operations to run EDIS is listed next. 

1. Log in on "bahama" and go to the directory which contains the EDIS knowledge bases. Only bahama 
has a valid NEXPERT license at this time. 

2. Run NEXPERT using the "nexpert &" command. 

3. Load the four EDIS knowledge bases kblllb.tkb, kblllctkb, planner.tkb, and qualitac.tkb in this or- 
der. 



Final Report July 1993 39 



Enhancements to the Engine Data Interpretation System 
(EDIS) 

4. "Volunteer M the value TRUE for data object USE_HEURISTICJRULES. Value only if you want to 
run heuristic rules. 

5. "Volunteer" the value TRUE for data object USE JPBM_DATA. Value only if you have and want to 
use PBM data. 

6. "Suggest" hypothesis LOAD_SAFE. 

7. Start NEXPERT knowledge processing with "Knowcess". 

8. When prompted enter the configuration directory. This may be absolute or relative to the current direc- 
tory. Always end with a V. 

9. You may want to watch progress in the Transcript window. Finally, the Session Control window will 
report that NEXPERT is done. Check out the object BEST_SCENARIO which contains information 
about the best diagnosis EDIS could find. 

10. Example 

In this section we will demonstrate the performance of EDIS with an example. The example case is test 
A1614 where a MCC Cooling leak was diagnosed by the SSME experts. Test A1613 was chosen as compari- 
son test and the following anomalies were reported. 



LPFP_DS_PR 


LOW 


MCCCLNT,DS_PR 


LOW 


MCC_CLNT_DS_TMP 


LOW 


LPFT_INLET_PR 


LOW 


LPFP_SPEED 


LOW 



Appendix A.5.1 contains a list of all measured parameter values in qualitative form. The anomalies listed 
above can be found there and all other parameter values are shown to be normal EDIS expects such a set of 
qualitative parameter values as input. EDIS, however, assigns values to only those parameters which are 
listed as " ASSOCIATE_PARAMETERS" (strange wording due to foreign graduate student) in the configu- 
ration file of any of the component types. In this set of files, parameters are identified by their standard names 
as recorded in the configuration files, such as LPFP_DS_PR. 

Appendix A.5.2 lists the larger set of numerical PBM data from both tests for a time slice of 10 seconds start- 
ing at 395 seconds into the test. These data are transformed into qualitative form and are read when PBM 
predictions are used to guide EDIS. They are identified by their location number in the A- ARRAY file. The 
original anomalies can be found there, too. For example, the lower than expected LPFP discharge pressure 
(LPFP_DS_PR) can be found at location 485 in file PBM_numeric_deviations, which indicates a 4.73% 
drop in pressure. _ 

10.1 Standard Operating Mode 

In standard operating mode EDIS iterates over the components in a manner which minimizes guessing. The 
fuel low controller is analyzed first because all its parameters are known. Thereafter, EDIS follows the net- 
work of component interconnections, analyzing one component at a time. Pipe splits and joins are avoided 
because they introduce a large degree of ambiguity. In the example, the next component analyzed is F101, the 



40 Final Report, July 1993 



Enhancements to the Engine Data Interpretation System 

(EDIS) 

pipe/duct from LPFP to HPFP, which contains the fuel flow meter. Then the HPFP is analyzed, etc. Appendix 
A. 5.3 lists the transcript of the NEXPERT session. The following faults are hypothesized in order. 



1. 


HPFP 


low efficiency 


2. 


HPOP 


low efficiency 


3. 


LPOP 


low efficiency 


4. 


LPFP 


low efficiency 


5. 


LPFP 


low efficiency 
behavior) 


6. 


MCC Cooling 


leak 



low efficiency (same fault mode as 4., but slightly different component 



Whether a fault is hypothesized at a component depends on whether one or more fault modes are consistent 
= with the measured and previously assumed parameter values and how likely the matching fault modes are. 

Appendix A.2.2 contains a listing of file faults which associates likelihoods with specific faults. A value of 
^ above 0.2 will normally direct EDIS into attempting to hypothesize the given fault mode. A value less than 

™ that indicates that the fault mode should only be considered after all "better " fault modes have proven unlike- 

ly or inconsistent. 

'— Behavior hypotheses with the smallest number of anomalous parameter values are preferred among compet- 

ing behavior hypotheses in the no-fault, i.e. normal, behavior mode. Fault modes are not differentiated in this 
™ manner. 

~~ 10,2 Using PBM Data 

^ In this mode of operation, EDIS prefers component behaviors whose parameter values agree with the values 

B predicted by the PBM. A transcript of EDIS executing with this option enabled is shown in Appendix A.5.5. 

EDIS does not operate significantly different from the standard case. The reason for the similarity is that the 
11 PBM predicts most parameter values to be normal and normal values are preferred by EDIS in its normal 

search mode. This, in fact, verifies the validity of the heuristics from which the heuristic evaluation function 

of EDIS was derived. 

m 10.3 Using Heuristic Rules 

sus EDIS does not readily find a solution when heuristic rules are used at the beginning of the session even though 

H the rules correctly identify the MCC Cooling leak as the cause of the observed anomalies. Too few parameters 

are known (only 3 out of 10) for the MCC Cooling duct, too many behaviors are possible (40 in this case), and 
zz too many of these behaviors (25) are instantiations of the LEAK fault mode. EDIS has no facilities to make an 

iJ informed choice between these behaviors and tries them in arbitrary order. 

„ 10.4 Using Heuristic Rules and PBM Data 

W Executing heuristic rules and using PBM data proved to be the most efficient way to solve the given diagnos- 

tic problem. EDIS identifies the correct fault hypothesis (MCC Cooling leak) at the start using its heuristic 
™ rules, then it identifies the correct MCC^Sbllng Savior by matching the 25 competing behaviors (see 

m above) against the PBM predictions, and then EDIS analyzes the remaining components to makesure that the 

proposed hypothesis is consistent with the data and the behavior constraints of all components. The heuristic 



Final Report July 1993 41 



Enhancements to the Engine Data Interpretation System 
(EDIS) 

which selects the most normal behavior for all the remaining components makes it possible for EDIS to select 
the "right" behavior for all remaining components on its first try. No backtracking at all needs to be per- 
formed. This is definitely an ideal and unexpected situation. The transcript of this session is shown in Appen- 
dix A.5.6. 

The leak behaviors hypothesized by the standard method and this method differ slightly. The former suggests 
a smaller leak than the latter, i.e. the standard method guesses V out to be NORMAL, while here V out is sup- 
posed to be LOW. Both have to assume that Vi n is HIGH. Since output pressure p out is LOW in both cases the 
fact that the LPFT produces lower than expected power (MechPWR = LOW) does not resolve the ambiguity 
either. The ambiguity illustrated by this example is inherent in EDIS because there are too few measurements 
available to uniquely identify the behavior of each component. Future versions of EDIS might be able to 
analyze and present equivalent cases such as these together. 

11. Known Limitations 

1. OPOV control to maintain power level is not modeled yet. 

2. The main combustion chamber (MCC) model assumes combustion at the optimal mixture ratio where 
any decrease or increase in LOX flow reduces combustion efficiency and output pressure. After study- 
ing the controller behavior in more detail we discovered that LOX flow is still used to control and main- 
tain power, i.e. MCC pressure. Therefore we conclude that the MCC must be operating on the slope 
instead of the plateau of the pressure curve and the implemented model is wrong. 

3. The interface to the anomaly detection system could not be implemented because of the immature de- 
velopmental status of the specification of the record formats for the anomalies and uncertainty about 
the interaction protocols between the anomaly detection modules and EDIS. 



4. Power balance data have to be transferred to the correct EDIS directory by hand. 

5. Anomalies cannot be distinguished by size. Only three qualitative values are available. 

11 Future Work 

1. Integrate EDIS with the PTDS and the Motif user interface. Use data classified by PTDS. 

2. Verify and refine the component models. Some of the models make assumptions which may not al- 
ways be true or may be oversimplified. For example, the pump models do not take the temperature 
increase of the pumped fuel or LOX into account. Larger than usual temperature increases may, how- 
ever, indicate pump efficiency problems. 

3. Test EDIS on more real cases. If necessary, add and/or modify component models. 

4. Improve the search process. A large amount of search can be avoided if scenarios are allowed to recon- 
vene after being split. In the current version, separate scenarios are maintained as long as scenarios 
differ in at least one parameter value. Scenarios are, however, equivalent if they predict the same fault 
(possibly none) and the parameter values at the "boundary" of the analyzed components have the same 
values. Thussome parameter values "inside" the analyzed ^components, i.e. within a component or at 
the interface between two analyzed components, may differ but the remaining search is identical for 
such a set of scenarios. They could be recombined into a single "aggregate-scenario." Savings of 
search time and storage space appear likely to be achievable by this approach. 



42 Final Report, July 1993 



u 



Enhancements to the Engine Data Interpretation System 

(EDIS) 

5. Design a general solution to the problem of differing classification scales. Currently, special methods 
are used to deal with the classification of parameter values at interfaces between components which 
operate at widely diverging operating points. For example, pressure deviations at the HPFP input can 
be analyzed with respect to either the input (low) pressure level or the output (high) pressure level. This 
is discussed in detail in the section on HIPUMP behavior. Again at the DIFFUSER, fuel flow is distrib- 
uted unevenly and flow rate deviations may be categorized against differing scales. The UN- 
EVEN JTHREE_SPLIT component model was developed to manage this case. It would be much 
better and lead to a more maintainable system if a general solution to this problem was implemented. 

6. Improve the heuristic evaluation function. For example, small failure effects which do not result in 
prima facie anomalies could be used to strengthen or weaken confidence in fault hypotheses. A small 
fuel leak, for example, most of the time causes a small increase in LOX flow and OPOV position as the 
controller is trying to maintain the requested power level. Both effects may be small enough not to be 
considered anomalies by themselves. When the diagnostic system is evaluating competing hypotheses 
and a fuel leak is proposed based on obvious anomalies, presence of such small scale effects could lead 
to increased confidence in a hypothesis. 

7. Investigate whether pre-start analysis results would facilitate diagnosis. The interview transcripts fre- 
quently mention expectations for engine behavior and measurement values based on information 
gained from analysis of engine pre-start behavior. It is not clear whether this information impacts only 
anomaly detection or could also assist in fault diagnosis performed by EDIS. 

8. Enlarge the knowledge base of heuristic rules used to identify likely faults and guide the qualitative 
reasoning system. 

9. Re-implement the diagnostic system in CLIPS in order to make it easier to incorporate it into the com- 
plete diagnostic systems. 

10. Complete the proof-of-concept fuzzy system and evaluate its performance relative to the purely quali- 
tative system. 

13. Conclusions 

The current version of EDIS contains models fgr aJJ major engine components, has a fully functional diagnos- 
tic reasoning module, and accepts suggestions generated by heuristic rules and by PBM data reduction. EDIS 
has not been extensively tested. All tests were done using a single MCC Cooling leak fault that occurred at 
test number A1614. During these tests we discovered a few small problems with our models which were due 
to the simplifications applied. It is to be expected that other cases will uncover additional modification re- 
quirements. We recommend a series of tests on a larger number of cases. 

EDIS is able to find common faults with current resource limitations and management but more difficult 
faults, i.e. unexpected and multiple faults, may exhaust the available time and memory resources, see below. 
Additional refinements to the search proc^de^bed in Section 12 above and enhancements to the resource 
management are necessary before EDIS is deployed and used on a day-to-day basis. 



The anomaly detection process is not necessarily exact because it depends on human judgement in a variety of 
ways. The current version of EDIS is not forgiving at all when confronted with a set of anomalies which is not 
consistent with expected component behavior modes. In some cases this problem may lead to the discarding 
of the correct solution. Crisp qualitative models can not efficiently deal with classification inconsistencies. 



Final Report July 1993 43 



Enhancements to the Engine Data Interpretation System 
(EDIS) 

We hope that the version of EDIS discussed in Section 8, enhanced with fuzzy classification and logic, will 
provide an effective and efficient remedy. 

CAVEAT EMPTOR: The implemented diagnostic system explores the space of all possible solutions, i.e. 
all possible behaviors of the SSME. It enumerates behaviors with and without faults; but with the currently 
supplied heuristic bias it prefers behaviors with a single fault. Note that the size of the search space grows 
exponentially with the number of parameters and thus with the number of components. It is therefore possible 
that the program will run out of memory or fail to give an answer within a reasonable time. Only the use of 
heuristics makes it possible to diagnose realistic anomalies. Without heuristics there would be no hope of 
finding a good solution. However, heuristics may fail and the system may propose a "wrong" diagnosis or 
none at all. Even if the system works correctly, its diagnosis may not identify the actual fault. From the sys- 
tem *s point of view a diagnosis is correct if it identifies the most likely fault given the available measurements 
and the supplied heuristics. Unfortunately this fault may yet be different from the actual fault. Note that bias 
due to knowledge limitations is a problem inherent to all machine and human reasoning. 

The algorithm implemented is a version of A* search. This type of search algorithm is guaranteed to find the 
best solution and to find it first but only if the heuristic evalua tio n function consistently underestimates the 
actual cost (or badness) of each evolving solution. We tried to use such a luhctibn at the beginning of the 
project and quickly discovered that the algorithm lacked focus on likely faults expected of an expert system. 
It tended to explore low likelihood areas of the search space because the evaluation function did not penalize 
these unlikely solutions enough, while making sure that even an unlikely solution would be found. The final 
version of the heuristic evaluation function is not guaranteed to underestimate the cost of evolving solutions 
and therefore might pass up the best solution in the first attempt. No possible solutions are totally disre- 
garded, however, they are just considered later in the search. The new algorithm draws broader conclusions 
from instances when an assumption cannot be justified , i.e. explained by a complete high quality scenario 
representing engine behavior. It will assume that the given assumption is bad and retract it after the first justi- 
fication attempt has failed, even if some other means of justifying it might actually succeed. The rational for 
this behavior is derived from the fact that the algorithm always attempts to complete the most likely justifica- 
tion first. The implemented algorithm therefore does not guarantee that that solution proposed first is the best 
one, but with reasonable heuristic information the most likely solutions will be generated before the less like- 
ly ones. As always, the meaning of "likely" depends on the heuristics, i.e. if the system is told that pumps fail 
more frequently than pipes it will prefer solutions that imply pump problems over those that imply pipe prob- 
lems. 

Pure qualitative models do not adequately model a system such as the SSME. Tie problem is that the same 
parameter value is interpreted differently depending on which component is analyzed. For example, the out- 
put pressure of the LPFP is actually identical to the input pressure of the HPFP (neglecting the duct between 
them for this example). A change in this pressure may, however, be considered significant, i.e. anomalous, 
when viewed in the context of the LPFP, and negligible when viewed in the context of the HPFP. The vastly 
different absolute values of pressure at the outputs of the LPFP and H PFP cause this discrepancy. The same 
size change will appear significant relative to the absolute pressure value at the LPFP and insignificant at the 
HPFP. A possible solution to this problem is to neglect changes in the HPFP input pressure values. We de- 
fined a separate "High-Pressure Pump" model in our system which implements this behavior. It appears 
reasonable to assume^^ much larger operating valued their output compared to 

their input would tend to "hide" deviations which are passed unchanged in size from input to output. Relative 
to the operating point the same size change will, of course, appear much smaller at the output. 



44 



Final Report, July 1993 



Enhancements to the Engine Data Interpretation System 

(EDIS) 

Another example is the confluence of flows of different magnitudes. TTie resultant flow may not be affected at 
all by a change in magnitude of a few percent of the smaller contributor. Rather than using specialized models 
for all these special cases, we conclude that it would be better if the model was aware of the difference in 
absolute magnitude. The purely qualitative model must then be extended with quantitative information. The 
fuzzy set theory-based system and the Order-of-Magnitude based system represent two attempts at coping 
with large variations in operating values. 

The computational complexity of the search for a consistent parameter value assignment, i.e. a scenario 
which explains the observer data and anomalies, is exponential . Memory requirements and computation time 
may grow excessively. In our implementation memory resource limits are the critical bound and it is entirely 
possible that process memory limits are reached. 

14. References 

[1] Johan de Kleer and John Seely Brown, "A Qualitative Physics Based on Confluences," in D. G. Bo- 
brow, (ed.), Qualitative Reasoning about Physical Systems, Cambridge, MA: MTT Press, 1985, pp. 
7-83. 

[2] Martin O. Hofmann, Thomas L. Cost, Michael Whitley, "Model-Based Diagnosis of the Space Shuttle 
Main Engine/' Artificial Intelligence for Engineering Design, Analysis and Manufacturing (AI EDAM), 
Vol. 6, No. 3, pp. 131-148. 

[3] Mark S. Fox, Norman Sadeh and Can Baykan, "Constrained Heuristic Search," Proceedings IJCAI-89, 
August 20-25, Detroit, Michigan, pp. 309-315. 

[4] A. K. Mackworth, "Consistency in Networks of Relations "Artificial Intelligence, Vol. 8, No. 1, 1977, 
pp. 99-118. 

[5] Elaine Rich and Kevin Knight, Artificial Intelligence, (2nd. ed.), New York: McGraw-Hill, 1991. 

[6] Hofmann, Martin O., "Model-Based Diagnosis Directed by Heuristic Search," Proceedings of the 
Ninth Conference on Artificial Intelligence for Applications, Orlando, Florida, March 1993, pp. 
197-203. 

[7] Kalagnanam J., Simon H. and Iwasaki Y. 1991, The Mathematical Bases for Qualitative Reasoning, 
IEEE Expert, 6(2), 11-19. 

[8] Didier Dubois and Henri Prade, Possibility Theory, New York: Plenum Press, 1988. 



Final Report July 1993 45 



Enhancements to the Engine Data Interpretation System 
(EDIS) 



APPENDIX 



46 Final Report, July 1993 



LJ 



B 



^ 



Type PIPE (File pipe) 



A.2 SSME Configuration Files 



A.2.1Component Files 



Type PIPE (File pipe) 



\F101.NAMEV="F101" 

\F10i.MEDIUM_INPUT\="LPFP" 

\F101.parameter_coupled_to_pin\="pout" 

\F101.parameter_coupled_to_\^n\="Vout" 

\F101.parameter_coupledjoJTin\="Tour 

\F101.MEDIUM_OUTP17A="HPFP" 

\F101.parameter_coupled_to_pout\="pin" 

\F101.parameter_coupledJo_Vout\= ,, Vin" 

\F101.parametcr_coupledjo_Tout\= , Tin M 

\F101.ASSOCIAIEJ?ARAM^^ 

\F101.GE3^RIC_PARAMErERS\= w pin,Tin,Vout" 

\F108.NAME\="F108" 

\F108.MEDIUM_INPUn="M103" 

\F108.parameter_coup]ed_to_pin\="pin" 

\F108.parameter_coupledjo_Vm\="VoutB" 

\F108.parameter_coupled_to_Tin\=*Tin" 

\Fi08.MEDIUM_OUTPUT\="OPB" 

VFlOS.parameter_coupled_to_jx>ut\="pin" 

\F108.parametcr_coupledJo_Vout\= n Vin" 

\F108.parameter_coup]edJoJTout\="Tin" 

\F108.ASSOCIATE_PARAMETERS\="Notknown" 

\F108.GENERIC_PARAMETERS\= w Notknown" 

\F110.NAME\= W F110" 

\FHO.MEDIUM_INPUT\="M103" 

\Fil0.parameter_coupled_to_pin\="pin" 

\F110.paramctcr_coupledJo_Vin\="VoutC" 

\F110.parametej coupled_to_TinWTin" 

\F110.MEDIUM"OUTPUT\="FPB" 

\F110.paramcter_coupled_to_pout\="pin" 

\F110.paramcter_coupledJo_Vout\= , *Vin" 

\Fll0.parameter_coupled_to_Tout\='Tin" 

\FllO.ASSOCIArE_PARAMFrERS\="Notknown" 

\F110.GENERIC_PARAMETERS\="Notknown" 

\F109.NAME\='T109" 

\F109.MEDIUM_INPIJA="MCC_COOL" 

\F109.parameter_coupled_to_pin\="pout" 

\F109.parametcr_coupledJo_Vin\="Vout" 

\F109,parameter_coupled_to Tin\="Tout" 

\F109.NfEDIUM_OUTPim="LPFT" 

\F109.parameter_coupled_tojx>ut\="pin" 

\F109.parameter_coupledJoJtoutVVin" 

\F109.parametcr_coupledJo_Tout\=*Tin" 

\F109.ASSOCIArej>ARAMETERS\="MC^^ 

\F109.GENERIC_PARAMETERS\="pin,Tin,pour 

\F190.NAME\="F19(T 

\F190.MEDIUM_INPUT\="LPFT , 

\F190.paramctcr_coupledJo_pin\="pout" 

\F190.parameter_coupled_to_\ r m\="Vout" 

\F190.parameterj;oupledJojrin\="Tout" 



262 Final Report, July 1993 

PRECEDING PAGE BLANK NOT FILMED 



\F190.MEDIUM_OUTPirA="Notknown" 

\F190.parametcr_coupIed_to_pout\="Notknown ,> 

\F1904>arameter_coupled_to_Vout\=^otknown" 

\F190.parameter_coupled_to_Tout\= w Notknown M 

\F190.ASSOCIAre_PARAl^TERS\=^otknown w 

\F190.GENERIC_PARAMETERS\= M Notknown w 

\O204.NAME\="O204" 

\O2Q4.MEDIUM_INPUTy= w M104 M 

\O204.parameter_coupIed_to^in\="pin" 

\0204.pararaetcr_coupIedJo_Vin\="Vout'* 

\O204.parameter_coupled to_Tin\="Tin" 

\O204.NffiDIUM_OUTPUT\="MOV" 

\O204.parameter_coupJed_to_pout\="pin" 

\O204.parameter_coupled_to_Vout\="Vin" 

\O204.parameter_coupled_to_Tout\='Tin" 

\02(M.ASSCK:iArE_PARAMETERS\= , 'Notknown" 

\O204.GENERIC_PARAMETERS\= ,, Notknown w 

\O203.NAME\="O203" 

\O203.MEDIUM_INPUT\="M104" 

\O203.parameter_coupled_to_pin\= M pin" 

\O203.parameter_coupIed_to_Mn\= M VoutB" 

\O203.parameter_coupledJoJTinVTiiT 

\O203.MEDIUM_OUTPUT\="LPOT , 

\O203.parameter_coupIed_to_pout\="pin" 

\O203.parameter_coupIed_to_Vout\="Vin" 

\O203.parameter_coupled_to_Tout\='Tin" 

\O203.ASSOCIATE_PARAMETERS\="Notknown" 

\O203.GENERIC_PARAMETERS\="Notknown M 

\O205.NAME\="O205 M 

\O205.MEDIUM_INPUT\="M104 M 

\O205.parameter_coupled_to_pin\=="pin" 

\O205.parameter_coupIedJo_Vin\="VoutC" 

\O205.parameter_coupledJoJTin\="TiiT 

\O205,MEDIUM_OUTPUT\="M101 M 

\O205.parameter_coupled_to_pout\="pin" 

\O205.parameter_coupledJo_Vout\="Vin" 

\O205.parameter_coupledJoJTout\="TiiT 

\O205.ASSOCIATE_PARAMETERS\="Notknown" 

\O205.GENERIC_PARAMETERS\="Notknown M 

\O206.NAME\="O206" 

\O206.MEDIUMJNPUI\="M101" 

\O206.parameter_coupled_to_pin\="pin M 

\O206.parameter_coup]ed_to_Vin\="VoutC" 

\O206.parameter_coupled_to_Tin\= , Tin" 

\O206.MEDIUM_OUTPim= M FPOV" 

\O206.parameter_coupIed_to_jx>iit\="pin" 

\O206.parameter_coupIedJoJ/out\='Vm" 

\O206.parameter_coupledJo Tout\="Tin" 

\O206.ASSOCIATE_PARAlvffiTERS\= w Notkiiown" 

\O206.GENERIC_PARAMETERS\="Notknown w 

\O201.NAME\="O201" 

\O201.MEDIUM_INPLn^= n LPOP ,, 

\O201.parameter_coupled_to_pin\="pout" 

\O201.parameter_coupIedJo_VIn\= M Vout M 



Final Report, July 1993 263 



\O201 .parameter_coupled to _Tin\="Tout M 

\O201.MEDIUM_OUTPUT\= M HPOP_PBP" 

\O201.parametcr_coupled_to_pout\="pin w 

\O201.pa^amctc^ - coupled_toJVbut\="\^n ,, 

\O201.parameter_coupledjoJTout\='Tin" 

\O201 .ASSOCIAIE_PARAMETERS\="LPOP_DS_PR" 

\O201.GENERIC_PARAMETERS\= w pin" 

\O190.NAME\="O190" 

\O190.MEDIUM_INPUT\=:"LPOT" 

\O190.parametcr_couplcd_toj)in\="pout" 

\O190.parametcr_couplcdJo_Vm\="Vout" 

\O190.parametcr_couplcdJojrin\= , Tout" 

\O190.MEDIUM_OUTPim= w Notknown" 

\O190.parameter_coupledJo^ut\=**Nbtknown w 

\O190.parametcr_coupled_to_Vout\= ,, Notknown" 

\O190.parametcr_coup]cd_to_Tout\="Notknown" 

\O190.ASSOCIATE PARAMETER3\=^otknown" 

\O190.GENERIC_PARAMETERS\= M Notknown" 

\F107.NAME\="F107" 

\F107.MEDIUM_INP17A="MIXER" 

\F107.parameter__coupled_to_pin\="pin" 

\F107.parameter_coupIed_to_Vin\="Vout" 

\F107.parameter_coupIed_to_Tin\='Tout" 

\Fl07.MEDIUM_OUTPtrn= H M103" 

\F107.paramctcr_coupIed_tojpout\= n pin" 

\F107.parameter_coupledJo_Vout\="W' 

\F107.parameter_coupled to Tout\="Tin" 

\F107.ASSOCIAre_PARANffiTERS\= w Notknowii" 

\F107.GENERIC_PARAMETERS\="Notknown M 



********** 



264 Final Report, July 1993 



Type COOLING (File cooling) 



w Type COOLING (File cooling) 

_ \MCC_COOL.NAME\="MCC COOL" 

= \MCC_COOL.MEDIUM_INPU ; A= M DIFFUSER" 

\MCC_COOL.parameter_couplcdjo_pin\= M pin" 

WlCC^COOL.parameter^coupledjo^in^^VoutC" 

\MCC_COOL.pa^amete^_coupledJo_Tin\= , Tm ,, 
™ \MCC_COOL.MEDIUM_OUTPUT\="F109" 

\MCC_COOL.parameter_coupled_to_pout\="pin" 
~~ WCC^OOL.paramete^coupledjo^Vout^'Vm" 

„ \MCC_COOL.paramcter_couplcd_toJToutVTin" 

\MCC_COOL.COOLS\="MCC" 

WCC^COOL.paramete^coupledJoJTsource^^Tout" 

\MCC_COOL.ASSOCIATE PARAMETERS\="MCC CLNT_DS_PR>ICC_CLNT DS TMP" 
"" \MCC_COOL.GENERIC_PARAMETERS\="pout,Tout w 

\NZL_COOL.NAME\="NZL COOL" 
M \NZL_COOL.MEDIUM_DWUT\=:"DIFFUSER" 

» \NZL_COOL.parameter_coupled_to_pin\="pin" 

\NZL_COOL.parameter_coup1ed_to_Vin\="VoutB" 
^ \NZL_COOL.parameter_coupIed toJTin\=*Tin w 

H \NZL_COOL.MEDIlM_OUTPUT\="MIXER" 

\NZL_COOL.parameter_coupled_tojx>ut\="pin" 
s \NZL_COOL.paramctcr_coupled_to_Vout\="VinB" 

■ VNZL^OOL.paramete^coupIedjoJTout^'TinB" 

— \NZL_COOL.COOLS\="NOZZLEl" 

\NZL_COOL.paramcter_coupled_to_Tsource\=*Tout" 
m \NZL - CCWDL.ASSOCIAre_PARAMETERS\= ,, Notkno\vn" 

5 \NZL_COOL,GENERIC_PARAMETERS\="Notknown" 

********** 



Final Report, July 1993 265 



Type VALVE (File valve) 

Type VALVE (File valve) 

\FPOV.NAME\="FPOV" 

\FPOV.MEDIUM_INPUT\= M O206" 

\FPOV.parametcr_coupled_to_pin\="pout" 

\FPOV.parameter_coupIed_to_Vin\= ,, Vout" 

\FPOV.paramcter_coupledJoJTin\= J Tout" 

\FPOV.MEDIUM_OXJTPUT\="FPB" 

\FPOV.paramctcrjCOiipled_tojx)iit\="pin_OX" 

\FPOV.pa^amete^_coupled_to_yout\= M Vln_OX ,, 

XFPOV.paramete^coupledJoJTout^'Tui^OX** 

\ftov.controlled_by\= m fuel_fl6w_ctrl" 

\FPOV.pa^ameter_coupledJo_commandedJX)sition\= w commandcd_position ,, 

\FPOV.ASSOCI£mj>ARAMETERS\='TTO^ 

^OV.GEOTRICJV^RANIETERSV^ 

\MFV.NAME\="MFV" 

\MFV.MEDIUM_INPim="HPFP" 

\MFV,parameter_coupled_to__pin\="pout" 

\MFV.paramctcr_coupIedJo_Vin\="Vout" 

\MFV.parameter_coupled_to_Tin\='Tout" 

\MFV.MEDIUM_OUTPim=' , DIFFUSER n 

\MFV.parameter_coupled_to_pout\="pin" 

\MFV.parameter_coupled_to_Vout\= w Mn" 

\MFV.parameter_coupled_to jrout\=*Tin" 

\MFV.CONTROLLED_BY\= M MFV_CTRL" 

\\{FV.parameter_coupled_to_commandcdj3osition\="commandedjx>sition" 

\MFV.ASSOCIAIEJVtflAMETERSV='TiPF^^^ 

\MFV.GENERIC_PARAMETERS\^ 

\CCV.N AME\= "CC V" 

\CCV.MEDIUM_INPim="DIFFUSER" 

\CCV.paramcter_coupled_to_pin\="pin" 

\CCV.parameter_coup]ed_to_Vin\="Vout" 

\CCV.parameterjx>upledJojrin\=Hm" 

\CCV.MEDIUM_OUTPUT\="MIXER M 

\CCV.parametcr_coupled_to_pout\="pin" 

\CCV.parameter_coupledJo_Vout\="Vin" 

\CCV.parameter_coupled_tojrout\='Tin" 

\CCV.CONTROLLED_BY\= >, CCV_CTRL >> 

\CCV.paramcter_coupled_tojcomniandcd_position\="conimanded_position" 

\CCV.ASSOCIA^E_PARAMETERS\= ,, CCV_POS^^ON,CCV_POSITION , ' 

\CCV.GENERIC_PARAMETERS\="position,commanded_position" 

\MOV.NAME\= w MOV" 

\MOV.MEDIUM_INPUT\="O204" 

\MOV.parameter_coupled_to_pin\="pout" 

\MOV.parameter_couplcd_to_Vin\="Vout" 

WOV.parametcr^couplcdJo^Tin^'Tout" 

\MOV.MEDIUM_OUTPim="MCC" 

\MOV.parametcr_coupled_to_pout\= M pin_OX w 

\MOV.parametcr_coupledjo_Vout\="Vin_OX" 

\MOV.parametcr coupled Jo Tout\="Tin_OX" 

^OV.CONTROLLED.BYX^'MOV^CTRL" 

\MOV.parameterjcoupledjo_conunandedjx>sition\= w commanded_position w 

\MOV.ASSOCIAm_PARAMEIERS\= M MOV_POSITION,MOV_POSITION" 

\MOV.GE>ffiRIC_PARAMETERS\="position,a)nimandedj)osition" 



2 <j£ Final Report, July 1993 



H 



\OPOV.NAME\="OPOV" 

\OPOV.MEDIUMJNPUTV'M10r 

\OPOV.parameter_coupled_to_pin\=="pin" 

VOPOV.parameter^coupledjo^in^^VoutB" 

\OPOV.parameter_coupIedJojrin\="Tin" 

\OPOV.MEDIUM m OUTPUT\="OPB" 

\OPOV.paramcter_coupledJojx)ut\="piii_OX" 

\OPOV.parameter_coupIed_to_Vout\="Vin_OX w 

\OPOV.parameter_coupIedJo Tout\='Tinj3X" 

\OPOV.CONTROLLED_BY\= H OPOV_CTRL" 

\OPOV.parameter_coupled_to commanded_position\="commanded_position" 

\OPOV.ASSCOArejPARA^^ 

\OPOVGENERIC_PARAMETERS\="position,commandedJ)osition ,, 



********** 



i 



■ 



Final Report, July 1993 267 



Type PUMP (File pump) 
Type PUMP (File pump) 



\LPFRNAME\="LPFP" 

\LPFP.MEDIUM_INPim= > 'FUEL M TANK" 

\LPFP.parametcr_coupIed_to_pin\="pout" 

\LPFRparameter_coupledJo_Vm\="Vout" 

\LPFRparametcr_coupled_to_Tin\='Tout" 

\LPFP.MEDIUM_OUTPUT\="F101 M 

\LPFP.paramctcr_coupled_to_pout\="pin" 

\LPFRparameter_coiipled jo_Vout\="VhT 

\LPFP.parametcr_coupledJojrout\='Tin w 

\LPFP.COUPLED_TO\="LPFT' 

\LPFRparameter_coupled_to_omega\="omega" 

\LPFP.parameter_coupledJo_MechPWR\= ,, MechPWR" 

\LPFP.parameter_coupledJoJTq\='Tq" 

\LPFP.ASSOCIAre w PARAMETERS\="ENG_FUEL_IN^ 

LET_T\fP,LPFP_DS_PR,LPFP_DS M TMP,LPFP_SPEEDl" 

\LPFRGENERIC_PARAMETERS\^*pin,Tin,poiit,Tout,omega M 

\LPORNAME\= M LPOP" 

\LPORMEDIUM_INPim="LOX - .TANK" 

\LPORparameter_coupled_to_pin\="pout" 

\LPORparamctcr_coup]edJo_Vin\="Vout" 

VLPORparametcr^coupledJojrin^'Tout" 

\LPORMEDIUM_OUTPUT\="O201" 

\LPORparameter_coupled_to_pout\="pin" 

\LPORparameter_coupledJo_Vout\="Vm" 

\LPORparameter_coupled_to Tout\="Tin" 

\LPORCOUPLED^TO\="LPOT w 

\LPORparamcter_coupled_to_omega\="omega" 

\LPORparamctcr_coupled_to_MechPWR\="MechPWR M 

\LPORparameter_coupled_to TqVTq" 

\LPORASSCCIXrej>ARAMETC^ 

LET_TMP,LPOP_DS_PR,LPOP_SPEEDl" 

\LPORGENERIC_PARAMETERS\="pin,Tin,pout,omega" 



********** 



268 Final Report, July 1993 



"" Type HiPUMP (File hipump) 

— Type HIPUMP (File hipump) 

WFRNAME\="HPFP" 

_ \HPFP.MEDIUM_INPim= , T101" 

\HPFRparameter_coupIed_to_pin\="pout" 

\HPFP.parameter_coupledJo_Viii\="Vout" 

\HPFP.parameter_coupled_to_ i .Tin\= , Tout" 

\HPFP.MEDIUM_OUTPUT\= M MFV" 

\HPFRparameter_coupled_to_pout\= "pin" 

\HPFRparamctcr_coupIedJo_VoutWm" 
_ XHPFP.paramctcr^coupledJo^Tout^'Tin" 

\HPFP.COUPLED_TO\="HPFT" 
i : \HPFRparameter_coupIed_to_omega\="omega" 

« \HPFP.parameter_coupled_to_MechPWR\="MechPWR" 

— \HPFRparameter_coupledJoJTq\="Tq" 

\HPFRASSOCIAIEJ>ARAMETERS\=^ TMP,HPFPjSPEEDr 

M \HPFP.GENERIC_PARAMETERS\="Vin,pout,Tout,omega" 

m \HPOPJ>BRNAME\="HPOP/PBP" 

\HPOP_PBP.MEDIUM_INPUT\="O201" 

\HPOP_PBRparameter"coupIedjoj)in\="pout" 
^ \HPOP_PBP.parameter_coupIed_to_Vin\="Vout" 

\HPOP_PBP.parameter"coupled_to_Tin\='Tout" 

\HPOP_PBRNDEDIUM_OUTPUT\="M104" 

\HPOP_PBRparameter_coupled_to_pout\="pin" 

— \HPOP_PBRparameter_coupled_to_Vout\="Vin" 
\HPOP_PBRparametcr_coupled_to_ToutWTin" 
\HPOPJ>BRCOUPLEDJTO\="HPOT" 

— \HPOP_PBRparameter_coupIed_toj>mega\="omega" 
\HPOP_PBRparamcter_couplcd_to_MechPWR\="MechPWR" 
\HPOPJ > BRparameter_coupIedjoJTq\="Tq" 
\HPOPJ>BRASSOCIXrej>ARAMETERS\="H^ 
\HPOP_PBRGENERIC PARAMETERS\="pout,Tout,omega" 



m 



m 



a 



fcf 



m Final Report, July 1993 269 



Type HYDRAULIC_TURBINE (File hturbine) 
Type HYDRAULIC_TURBINE (File hturbine) 



\LPOT.NAME\="LPOT' 

VLPOT.MEDIUMJNPUT\="O203" 

\LPOT.parameter_coupled_to_pin\="pout" 

\LPOT.parameter_coupled_to_Vin\="Vout" 

\LPOT.parameter_coupled_to_Tin\='Tout" 

\LPOT.MEDIUM_OUTPUT\="O190" 

\LPOT.parameter_coupIed_to_pout\="pin" 

\LPOT.parameter_coupled_to_Vout\="Vin" 

\LPOT.paramcter_coupIed_to_Tout\='Tin" 

\LPOT.COUPLED_TO\="LPOP" 

\LPOT.paramctcr_coupled_to_omega\="omega" 

\LPOT.parameter_coupled_to_MechPWR\="MechPWR" 

\LPOT.pararacter_coupled_to_Tq\=Tq" 

\LPOT.ASSCX:iArE_PARAMETERS\="LPOP_SPEEDl" 

\LPOT.GENERIC_PARAMETERS\="omega" 



********** 



270 Final Report, July 1993 



Type GASJTURBINE (File gturbine) 

Type GASJTJRBINE (File gturbine) 

\LPFT.NAMEV="LPFT" 

\LPFT.MEDIUM_INPUT\="F109" 

\LPFT.parameter_coupIed_to_pin\="pout" 

^PFT.paramete^coupIedJo^nV^Vout" 

\LPFT.parameter_coupledJojnn\="Tout" 

\LPFT.MEDIUM_OUTPim=T190 M 

\LPFT.parameter_coupled_to_pout\=="pin" 

^PFT.parametc^cxjupledJoJ/out^'Vin" 

\LPFLparameter_coupled Jo JTout\= 'Tin" 

\LPFT.COUPLEDJTO\="LPFP" 

\LPFT.parameter_coupled_to_omega\="omega" 

\LPFT.parameter_coupled_to_MechPWR\= M MechPWR H 

\LPFT.paramcter_coupIedJo_Tq\="Tq M 

\LPFT.ASSCCIAre_PARAMETERS\=="LPFT_INLEr_PR,LPFP SPEED1" 

\LPFT.GENERIC_PARAMETERS\="pin ) omega tt 

\HPFT.NAME\="HPFT" 

\HPFT.MEDIUMJNPUT\='TPB" 

\HPFT.parameter_couplcd_to_pin\="pout" 

\HPFT.parameter_coupled_to_Vin\="Vour 

\HPFT.parameterj;oiipledJojrm\="Tout" 

\HPFT.MEDIUM_OUTPUT\="HGM" 

\HPFT.parameter_coupIed_to_pout\="pin" 

\HPFT.parameter_coupledJo_Vout\="Vin" 

\HPFT.parametcr_coupled_to_Tout\=*Tin" 

\HPFTCOUPLED_TO\="HPFP" 

\HPFT.parameter_coupled_to_omega\="omega" 

\HPFT.parameter_coupled_lo_MechPWR\= >, MechPWR w 

VHPFT.parameter^oupledJoJTqVTq" 

\HPFr.ASSOCDOE_PARAMETERS\= M FPB_PC,HPFr_DS_T\IPl,HPFP_SPEEDl w 

\HPFT.GENERIC_PARAMETERS\="pin,Tout,omega" 

\HPOT.NAME\='TO>OT" 

\HPOT.MEDIUM_INPUT\="OPB" 

\HPOT.parameter_coupled_to_pin\="pout" 

\HPOT.parameter_coupled_to_Vin\="Vout" 

XHPOT.parametc^coupIedJoJTin^'Tout" 

\HPOT.MEDIUM.OUTPUT\= M HGM M 

\HPOT.parameter_coupled_tojx)ut\="pin" 

\HPOT.parameter_coupIed_to_Vout\= ,, VinB w 

MIPOT.paramcter^coupledJo^out^'TinB" 

\HPOT.COUPLEDJTO\="HPOP_PBP" 

\HPOT.parameter_coupled_to_omega\= "omega" 

\HPOT.parametcr_coupledJo_MechPWR\="MechPWR" 

\HPOT.parameter_coupled_to_Tq\="Tq" 

\HPOT.ASSCOXrej>ARAN^^ 

VHPOT.GENERI^PARAMETERS^^inTout^mega" 

********** 



Final Report, July 1993 271 



Type PREJBURNER (File pbumer) 
Type PREJBURNER (File pburner) 



\FPB.NAME\="FPB" 

\FPB:FUELJN\= ,, Fll(r 

\FPB.parameter_coupIed_to_pin\="pour 

\FPB.paramcter_couplcd_to_Vin\= >, Vout" 

^B.parameter^coupledJoJTin^'Tout" 

^B.GAsjxm^Tnrf" 

\FPB.parameter_coupledJojx>ut\="pin" 

^Fra.parameter_coupIedjo_Vout\= w Vin w 

\FPB.paramcter_coupled_tojrout\='Tin" 

\FPB.OX_IN\= w FPOV" 

\FPB.parameter_couplcdJo_pin_OX\="pout" 

\FPB.parameter_coupIedJo_yin_OX\="Vout" 

\FPB.paramcter_couplcdJoJTinjOX\= , Tout" 

\FPB.ASSOCIATE_PARAMETERS\="FPB_PC" 

\FPB.GENERIC M PARAMETERS\="pour 

\OPB.NAME\="OPB" 

\OPB.FUELJN\="F108" 

\OPB.parameter_coupled_to_pin\="pout" 

\OPB.parameter - coupled_to_Vin\="Vout" 

\OPB.parameter_coupkdJoJTin\='Tout" 

\OPB.GAS_OUT\="HPOT" " 

\OPB.parameter_coupled_to_pout\="pin" 

\OPB.parameterj;oiipledjoJVout\="Vin" 

\OPB.parameter_coup1ed_toJTout\="Tin" 

\OPB.OXj^=' r OPOV" 

\OPB.paramcter_coup1ed_to_pin_OX\= M pour 

\OPB.parameter_coupled_to_V5n_OX\="Vout" 

VOPB.parameter^oupledJoJTinjO^s'Tout" 

\OPBASSOCLffE PARAME , rERS\= , *OPB_PC ,, 

\OPB.GENERIC_PARAMETERS\="pout" 



********** 



2J2 Final Report, July 1993 



Type MAIN_BURNER (File mburner) 
Type MA1N_BURNER (File mbumer) 



\MCC.NAME\="MCC" 

\MCC.FUELJN\="HGM" 

\MCC.parameter_coupled_to_pin\="pin M 

\MCC.parameter_coupledJo_\ r in\="Vout" 

\MCC.parametcr_coupled_to_Tin\= , Tout" 

\MCC.GAS_Oim="NOZZLEl" 

\MCC.parameter_coupled_tojx>iit\="pin n 

\MCC.parameter_couplcd_to_Vout\= H Vm" 

\MCC.parameter_coupledJoJToutB\=Hm" 

\MCC.COOLED_BY\="MCC.COOL" 

\MCC.parameter_coupled_to Tout\="Tsource " 

\MCC.OX_IN\= M MOV" 

\MCC.parameter_couplcd_to_pin_pX\="pout M 

\MCC.parameter"coupled_to_Vin_OX\= w Vout" 

\MCC.parameter_couplcd_to_Tin_OX\= , Tout" 

\MCC.ASSOCIATEJ > ARANffiTERS\="MCC_PC w 

\MCC.GENERIC_PARAMETERS\="pout" 



********** 



Final Report, July 1993 273 



Type CONTROLLER_CONST (File ctrlfuel) 



Type CONTROLLER_CONST (File ctrlfueO 

\FUELJ^OW_CTRL.NAMEV="FXJEL_FLOW CTRL" 

\FUEL_FLOw"cTRL.MEASURES_AA="F10r 

\FUELJT J OW~CTRL.parameter_coupled_to_Vin\= , 'Vout" 

\FUEL_FLOW_CTRL.CONTROLS\="FPOV" 

\FTJEL_FLOWjCTRL.parametcr_coupIed_to_conimanded_position\=''coinmanded_position'' 

\FUEL_FIX>W_CTRL^SSOCIAre_PARAMETERS\="FUEL_FLOW,FPOV_POSrnON^ 

\FUELJTX>W_CnU..GENERIC_PARAMETERS\="Vm,commanded_position" 

\MFV_Cim-.NAME\="MFV_CTRL" 

\MFV_CTRL.MEASURES_AT\="Notknown" 

\MFV_CTRL.parameter_couplcd_to_\Tn\="Notknown'' 

\NIFV_CTRL.CONTROLS\= , 'MFV" 

\MFVjCTRL.parameter_couplcd_to_coinmanded_position\="coininanded_positioii" 

\MFV CTRL.ASSOCIATE_PARAMETERS\="Notknown" 

\MFV~CTRL.GENERIC_PARAMETERS\="Notknown" 

\CCV_CTRL.NAME\="CCV_CTRL" 

\CCV_CTRL.MEASURES_AT\="Notknown" 

\CCV_CTRL.parameter coupIed_to_Vin\= "Notknown" 

\CCV_CTRL.CONTROLS\="CCV"~ 

\CCV_CTRL.parameter_coupkd_to_coininanded_position\="commanded_position" 

\CCV_CTRLJlSSOCIArE_PARAMETERS\="Notknown" 

\CCV_CTRL.GENERIC_PARAMETERS\= ,, Notknown" 

\MOV_Cr^RL.NA^ffi\= B MOV_CTRL , ' 

\MOV_CTRL.MEASURES_AA="Notknowii" 

\MOV_CTRL.parameter_coiipled to_Vin\="Notknown" 

\MOV_CTRL.CONTROLS\="m6V" 

\MOV_CTRL.parametcr_coupled_to_commanded_position\="commanded_position" 

\MOV_CTRL.ASSOCIATE PARAMETERS\="Notknown" 

\MOV_CTRL.GENERIC_PARAMETERS\= ,, Notknowri" 

\OPOV_CTRL.NAME\="OPOV CTRL" 

\OPOV_CTRL.MEASURES_AT\="Notknown" 

\OPOV_CTRL.parameter_coupled_to_\^n\="Notknown" 

\OPOV_CTRL.CONTROLS\="OPOV" 

\OPOV_CTRL.parameter_coupIed to_commanded_position\="commandcd_position" 

\OPOV_CTRL.ASSOCIAIE_PARAMETERS\=''Notknown" 

\OPOV_CTRL.GENERIC_PARAMETERS\="Notknown" 

* * * * ♦ * * * * * 



274 Final Report, July 1993 



Type TWO_SPLIT (File twospiit) 
Type TWO_SPLIT (File twospiit) 



\M103.NAME\="M103 M 

\M103JMEDIUM_IN\= M F107" 

\M103.parameterjcoupIed_toj>in\="pout" 

\M103.parameterjcoupIed_to_\^n\=' , Vout" 

\M103.parameter_coupled_to Tin\="Tout" 

\M103.MEDIUM_OUTAWTi08" 

\M103.parameter_coupIed_tojpoutA\="pin" 

\M103.parameter_coupled_to_VoutA\="Vm" 

WllOS.parameter^coupledJo^outA^Tin" 

\M103.MEDIUM_OUTB\= M F110" 

\M103.parameter_coupled_to_poutB\= n pin" 

\M103.parameter_coupled_to_VoutB\= M Vm" 

\M103.parameter_coupled_to ToutBWTin" 

\M103.ASSOCIAIE_PARANffiTERS\= M Notknown" 

\M103.GENERIC_PARAMETERS\= rt Notknown w 

\M101.NAME\="M101" 

\M101.MEDIUM_IN\="O205 w 

\M101.parameter_coupled_to_pin\="pout" 

\M101.parameterjcoupled_to_Vm\= M Vout" 

\M101.parameter_coupIed_to Tin\="Tout" 

\M101.MEDIUM_OUTA\="(5POV M 

^lOl.parameter^coupIed^to^poutA^^pin" 

\M101,parameter_couplcdJo_VoutA\="Vin" 

\M101.parameter_coupIed_toJToutA\='Tin" 

\M101.MEDIUM_OUTB\= M O206" 

\M101.parameter_coupled_to_poutB\="pin" 

\M101.parameter_coupledJo_VoutB\="Vin w 

\M101.parameter_coupledjoJToutB\='Tin" 

\M101ASSOCIA^E_PARAMETERS\= ,, Notknown ,, 

\M101 .GENERIC_PARAMETERS\="Notknown" 
********** 



Final Report, July 1993 275 



Type THREE_SPLIT (File trisplit) 
Type THREE_SPLIT (File trisplit) 



\M104.NAME\="M104" 

\M104.MEDIUM_IN\="HPOP_PBP" 

\M104.parameter_coupled_to_pin\="pout" 

\M104.parameter_coupledJo_Vin\= ,, Vout" 

XMKM.paramctc^coupledjtojrinV^out" 

\M104.MEDIUM_OUTA\="O204" 

\M104.parameter_coupled_to_poutA\="pin" 

\M104.parameter_coupledJo_VoutA\= ,5 \^n" 

NMKM.parameterjcoupled^toJToutA^'Tin" 

WKM.MEDIUM^OUTBW'OZOS" 

\M104.paramcter_coupled_to_poutB\= w pin" 

\M104.parametcr_coiipled_to_VoutB\= M Vin" 

\M104.pa^amete^_coupledJo_ToutB\= , Tin ,l 

\M104.MEDIUM_OUTC\= w O205" 

\M104.parameter_coupled_to_jx>utC\="pin" 

\M104.paramcter_coupledjo_VoutC\= M Vin" 

\M104.parameter_coi2pledJo_ToutC\="Tin" 

\Mi04ASSOCIAre_PARAMETERS\=' , HPOP_DS_PR,HPOPJDSJ^MP ,, 

\M104.GENERIC_PARAMETERS\= M pin,Tin" 



********** 



276 Final Report, July 1993 



Type UNEVENJHREE_SPLIT (File utrisplit) 



Type UNEVEN_THREE_SPUT (File utrisplit) 



\DIFFUSER.NAME\="DIEFUSER" 

\DIFFUSER.MEDIUM_IN\="MFV" 

\DIFFUSER.parameter_coup]ed_tojpin\="pout" 

\DIFFUSER.parameter_coup]ed_to_Vin\="Vout" 

\DIFFUSER.parameter_coupledJojrin\= , Tout w 

\DIFFUSER.MEDIUM_OUTA\= "CCV" 

\DIFFUSER.paiameter_coupled_to_poutA\="pin" 

\DIFFUSER.parameter_coupled_to_VoutA\="Vin" 

\DIFFUSER.parameter_coupledJojroutA\= , Tin" 

\DIFFUSER.MEDIUM_OUTB\= w NZL__CCMDL" 

\DIFFUSER.parametcr_coupled_to_poutB\="pin M 

\DIFFUSER.parametcr_coupled_to_VoutB\= ,t Vin" 

\DrFFUSER.parameter_coupled_to_ToutB\='Tin M 

\DIFFUSER.MEDIUM_OUTC\="MCC_COOL" 

\DIFFUSER.parameter_coupIedJo_poutC\="pin" 

\DIFFUSER.pa^amete^_coupIedJo_VoutC\="Vin ,, 

\DIFFUSER.parameter - coupIcd_to_ToutC\= , Tin" 

\DIFFUSER.ASSOCIATE_PARAMETERS\="Notkiiowii M 

\DIFRJSER.GENERIC_PARAMETERS\="Notk^owIl' , 

********** 



Final Report, July 1993 277 



Type TWO_JOIN (File twojoin) 
Type TWO_JOIN (File twojoin) 



\MKER.NAME\= ,, MIXER" 

\MKER.MEDIUM_INA\="CCV" 

\MIXER.parameter_coupled_toj?inA\= w pout" 

\MKER.parametcr__coupledJo_VinA\= ,, Vout" 

\MKER.parameter_coupled_to TinA\="Tout" 

\MKER.MEDIUM_OUT\= > T107 M 

\MIXER.parameter_coupled_to_poutA\="pin" 

\MIXER.parameter_coupled_to_Vout\="Vin" 

\MKER.parameter_coupledJojrout\="Tin M 

\MKER.MEDIUM_INB\= ,, NZL_COOL" 

\MIXER.parameter_couplcd_toj>inB\="pout" 

\NfIXER.parameter_couplcd_to_VinB\="Vout" 

\MIXER.parameter_coxipIed_tojrinB\="Tout" 

\MIXER.ASSOCIATE_PARANffiTERS\="Notknown" 

\MIXER.GENERIC PARAMETERS\= w Notknown" 

\HGM.NAME\="HGM" 

\HGM.MEDIUMJNA\="HPFT" 

\HGM.parameter_coupled_to_pinA\="pout" 

\HGM.parameter_coupled_to_VinA\="Vout" 

\HGM.parameter_coupled_toJTinA\="Tout" 

\HGM.MEDIUM_OUT\= w MCC" 

\HGM.parameter_coupled_to_poiitA\="piiT 

\HGM.parameter_coupledJo_Vout\= ,, Vin" 

\HGM.parameter_coupledJoJTout\="Tin" 

\HGMMEDIUMJNBVTJPOT" 

\HGM.parameter_couplcd_to_pinB\= M pout" 

\HGM.parameter_coupledJo_VmB\="Vout" 

\HGM.parameter_coupledJoJTinB\="Tout" 

\HGM.ASSOCIATE PARAMETERS\="HPFT_DS_T^IPl,HPOT_DS_'IMPr , 

\HGM.GENERIC_PARAMETERS\= , Tin,TinB ,, 



278 Final Report, July 1993 



Type NOZZLE (File nozzle) 
Type NOZZLE (File nozzle) 



\NOZZLEl^TAME\="NOZZLEl" 

\NOZZLEl.MEDIUM_INPUT\="MCC" 

\NOZZLEl.parameter_coupled_to_pin\="pout" 

\NOZZLEl.paramcter_coupled_to_Vin\="Vout" 

\NOZZLEl.parameter_coupled_to_Tin\='Tout" 

\NOZZLE1.COOLED_BY\="NZL_COOL" 

\NOZZLEl.parametcr coupled Jo_Tout\="Tsource" 

\NOZZLEl^VSSOCWaE_PARAMETERS\="MCC_PC" 

\NOZZLEl.GENERIC_PARAMETERS\="pin" 



********** 



Final Report, July 1993 279 



Type TANK (File tank) 

Type TANK (File tank) 

\FUEL_TANK.NAME\="FUEL_TANK" 

\FUEL_TANK.MEDIUM_OUTPUT\="LPFP" 

VFUEL_TANKparameter_coupled_tojx)ut\="pin" 

\FUEL_TANKparameter_coupIed_to_Vout\="\ln" 

VFUEL_TANK.parameter_coupled_to_Tout\= , Tin" 

\FUEL TANKASS(XI/aEJ > ARAMETERS\='WGJ=T^ 

\FUEL~TANK.GENERIC_PARAMErERS\=''pout,Tout'' 

\LOX_TANK.NAME\="LOX_TANK" 

\LOX_TANICMEDIUM_OUTPim="LPOP" 

\LOX_TANK.parameter_couplcd_to_pout\="pin" 

\LOX_TANK.parameter_coupled_to_Vout\="Vm" 

\LOX_TANK.parametcr_coupled to~Tout\='Tin" 

\LOXJTANKJ^SOCIArej>ARAMETERS\="ENGJD^^^ 

\LOX_TANK.GENERIC_PARAMETERS\="pout,Tout" 

********** 



280 Final Report, July 1993 



Class TERMINAL (File terminal) 



Class TERMINAL (File terminal) 



\FUEL_TANK.NAME\="FUEL_TANK" 

\F190.NAME\="F190" 

\LOX_TANK.NAME\="LOX_TANK" 

\O190TNAME\="O190" 

\MFV_CTRL.NAME\="MFV_CTRL" 

\CCV_CTRL.NAME\="CCV_]CTRL" 

\MOV CTRLJSrAME\="MOV_CTRL" 

\OPOV_CTRL.NAME\="OPOV_CTRL" 

********** 



Final Report, July 1993 281 



Fault Mode Likelihoods 



A.2.2Fault Mode Likelihoods 

Fault Mode Likelihoods 

\pipeJeakxomponent\="PIPE" 

\pipeJeak.fault\="LEAK" 

\pipeJeak.probabi1ity\='Ul" 

\pipe_obstnictionxomponent\="PIPE" 

\pipe_obstruction.fault\= w OBSTRUCTION" 

\pipe_obstruction.probabiIity\="0. 12" 

\coolingJeakxomponent\="COOLING" 

\cooling_leak.fauIt\="LEAK" 

\coolingjeak.probability\="0.2" 

\cooling_obstnictionxomponent\="COOLING" 

\cooling_obstruction.fault\="OBSTRUCTION" 

\cooling_obstruction.probability\="0.1" 

\pump_impeIIer_problcm.component\="PUMP" 

\pumpJmpeIIer_problcm,fault\="IMPELLER_PROBLEM" 

\puTnp_impcllcr_prob]eni.probabi!ity\="0.15" 

\pumpJow_efficiencyxomponentY="PUMP" 

\pumpJow_efficiency.fault\= , XOW_EFnCIENCY" 

\pumpjow_efficiency.probability\="0.3" 

\hipump_impeller_problem.component\="HIPUMP" 

\hipumpjmpeller_problem.fault\="IMPELLER_PROBLEM" 

\hipximp_impeller_problem.probability\="0.25" 

\hipumpJow_efficiencyxomponent\="HIPUMP" 

\hipump_low_efficiency.fault\="LOW_EFnCIENCY" 

\hipump_low_efficiency.probability\="0.4" 

\valve_blockage.component\="VALVE" 

\valve_blockage.fauIt\="VALVE_BLOCKAGE" 

\valve_blockage.probability\="0.1" 

\valvejservo_fauItxomponent\="VALVE" 

\valve_servoJauIt.fault\= w VALVE_SERVO_FAULT" 

\valve_servo_fault.probability\="0.08" 

\gas_tijrbineJow_efiBciencyxomponentV M GASjnJRBINE" 

\gas_turbmeJow_efficiency.fault\= w LOW_EFFICIENCY" 

\gas_turbinejow_efficiency.probability\="0.25" 

\hydraulicjurbinejmpellerjroblemxompo^^ 

\hyd^auIic_turbineJmpeIle^^roblem.fauIt\= w IMPELLER_PROBLE^^ , 

\hydraulic_turbine_impeller_problem.probability\=:"0.12" 

\controIIer_faultxomponent\="CONTROLLER_CONST' 

\controllerJault.fault\=XONTROLLER_FAULT" 

\controIIcrJault.probabiIity\= > U08" 
********** 



282 Final Report, July 1993 



heuristic-rules.tkb 



A.3 Heuristic Rules File 



heuristic-rules.tkb 



(@VERSION= 020) 

(©PROPERTY* component_classes ©TYPE-String;) 

(©CLASS- SSME OONRGURAT10N 
(©SUBCLASSES- 

fuel side 
lox Side 
) 
) 

(©CLASS- FUEL SIDE 
) 

(©CLASS- LOX SIDE 
) 

(©OBJECT- CONTROL HEURISTICS 
(©PROPERT1ES- 

current_behavior 
oomponent^ciasses 
booMemp " 
temporary 
) 
) 

(©OBJECT- FUEL SIDE RLE 
(©CLASSES'- ~ 

LOAD_CONTROL 
) 
{©PROPERTIES- 

class_name 

file_exist$ 

filename 

NAME 

retrieve 
) 
) 

(©OBJECT- LOX SIDE FILE 
{©CLASSES* 

LOAD CONTROL 
) 
(©PROPERTIES* 

class name 

fileexists 

ffle~name 

NAME 

retrieve 
) 
> 

(©OBJECT- AbemethyJ 01 108 Fuel Leak 
) " " 

(©OBJECT- Randy Hurt Fuel leak 
) 

(©OBJECT- Randy Hurt_HPFP_efficiency low 
) 



(©OBJECT- Randy_Hurt_MCC_Cooling_leak 

(©OBJECT- Ran<ty_Hurt.NZL_Cooling Jeak 

(©OBJECT- Randy Hurt HPOT efficiency low 
) 

(©OBJECT- SUGGESTIONJTYPES 
(©PROPERT1ES- 

Value ©TYPE-Boolean; 

) 
) 

(©SLOT- FUEL SIDE RLE.dass name 
{©SOURCES- 

(RunTimeValue fPJEL_SIDE")) 

) 

(©SLOT- FUEL SIDE RLE.NAME 
(©SOURCES- " 

(RunTimeValue puel side")} 



Final Report, July 1993 283 



heuristic-rules.tkb 



) 

(©SLOT* LOX SIDE RLExtass name 

(©SOURCES- " 

{RunTSmeValue fLOX SIDE")) 

) 
) 

(©SLOT- LOX SIDE FILENAME 

(©SOURCES- " 

(RunTimeValue ftox skJe"}) 

) 
) 

(©RULE- fueUine_leak 
©INFCAT-1; 

©COMMENT$*"componeri^name can alto hoW a pair (system taction class.component class), sae below for an example. \ 
Another set of files must be read to assign components to each system section"; 
(©LHS- 

(Is (<[SCENARJO BEHAVIORS|>.comp name) rHPOT)) 
0* (<|SCENAR10~BEHA\/IORSi>Tout) " ("HIGH")) 
(Is («|SCENAR10* BEHAVIORS|».comp name) CHPFT)) 

(Is («|SCENARIO~BEHAVIORSI»Tou^ TNORMAL","HK3H^) 
(Is <<« SCENARI5 BEHAVIORSi>».comp name) fHPFP")) 
(Is («<SCENARIO"BEHAVIORS>».pouQ" ("LOW")) 
(IsNot («< SCENAR»0"BEHAVIORSi>».omega) fLOW)) 
{Is (««1SCENAR(5 BEHAV10RSi»».comp name) fMCC")} 
(l« (««|SCEr4ARIO"BEHAV10RS|»».pouD " f NORMAL")) 
) 

(©HYPO- EXPERT HEURISTIC RULE) 
(@RHS- 

(CreateObject (Abemethy_101 106 Fuel Leak) a EXPERT RULE SUGGESTIONS^ 

{HEURISTIC SUGGESTIONSP) 

"(Let (Abemethy 101 108 Fuel Leak. component name) ("FUEL SIDE,DUCT» 
(Let (AbemethyJ0l"loe"FuerLeak.suggestedJault) (TEAiq) 
(Let (Abernethy~1 01 108 Fuel Leak. suggestion" type) ("component and_fautt_type")) 
) 
) 

{@RULE= fuel line Ieak2 
©INFCAT-0; 

@COMMENTS-"SmaJl-scale effects: OPOV position H, LPOP Vin H"; 
(@LHS= 

(Is (<|SCENARIO BEHAVIORS|>.comp_name) CHPFP")) 
(Is (<|SCENARiO"BEHAVIORS!>.pou0 fLOW)) 
(Is («|SCENARIO BEHAVIORSj».wmp name) ("HPOT")) 

(Is {<<|SCENARIO~BEHAVIORS|>>.Tout} "CHIGH")) 
) 

(©HYPO- EXPERT HEURISTIC RULE) 
(©RHS- 

(CreateObject (Randy Hurt Fuel leak) ((EXPERT RULE SUGGESTIONS! A 

(HEURISTIC SUGGESTIONS])) - - - - 

"(Let (Randy Hurt Fuel leak .component name) ("FUEL SiDE,DUCT)) 
(Let (RandylHurt'FuelJeak. suggested Jauft) CLEAfO) 
(Let (Randy Hurt Fuel leak. suggestion type) ("component and fault type")) 
) 
) 

(©RULE* HPFP efficiency low 
@INFCAT=0" 

@COMMENTS«"Small-scale effects: HPFP pout H, HPOTTout H"; 
(©LHS- 

(Is (<jSCENARlO BEHAVIORS(>.comp name) fFPOV"}) 

(Is (<|SCENARIO"BEHAVIORS|>.pos(tior)) ("HIGH")) 

(Is («[SCENARIO BEHAVIORS|».comp name) CHPFT}) 

(Is («lSCENAR»0~BEHAVIORS|».Tout) "fHiGHl) 

0« («<|SCENARl5 BEHAV!ORS|»>,comp name) ("FPB")) 

(Is (<«|SCENARIOlBEHAV10RS|>».pout) " fHlGH*)) 

(©HYPO- EXPERT HEURISTIC RULE) 
(@RHS= 

(CreateObject (Randy Hurt HPFP efficiency low) QEXPERT RULE SUGGESTIONS^ 

)HEURlSTiC_SUGGEST10NS|)) " ~ " " 

(Let (Randy Hurt HPFP efficiency low.component name) fHPFP")) 
(Let (Randy"Hurt*HPFP"efficiencyJow.suggestedJauR) fLOW EFFICIENCY*}) 
(Let (Randy "Hurt^HPFP efficiency low.suggestion"type)rspeciflc component)) 
) 
} 

(©RULE- MCC Cooling leak 
©INFCAT-0; 

©COMMENTS*"Small-scale effects: LPOP Vin H, MOV position H, OPOV position H, HPFP pout L"; 
(©LHS- 

(1s (<|SCENAR10 BEHAVIORSI>.comp_name) ("LPFP")) 

(Is (<jSCENAR!0~BEHAVJORS|>.ornefla) ("LOW")) 



(Is (« 
(Is (« 



SCENAR10_BEHAVJORS 
SCENARIO BEHAVIORS 



SCENARIO.BEHAVtORS ».pouQ ("LOW")) 



».comp name) fMCC COOL")) 

Tout)TLOVO) 



(Is («£ 

) 

(©HYPO- EXPERT HEURISTIC RULE) 

(@RHS- 

(CreateObject (Randy_Hurt_MCC.Cooling_leak) (|EXPERT_RULE_SUGGEST10NS1A 

|HEURISTiC_SUGGESTIONS))) 

{Let (Randy_Hurt_MCC_Cooling_leak.component_name) ("MCC_COOL")) 



284 Final Report, July 1993 



heuristic-rules.tkb 



(Let (RarK*y_h4urt_MOC_Coolir>g_.leak.»ua9est»d_teuft) H-EAK")) 

(Let (Randy Hurt MCC Cool ingjeak. suggestion type) fspecHtc component")) 
) 
) 

(@RULE= Nozzle Cooling leak 
©INFCAT-O;" 

@OOMMENTS»"Small-scale effects: HPFTTout H, OPOV position H, NZL COOL Vtn H, MCC COOL Vm L"; 
(@LHS- 

(ls (<|SCENARJO BEHAVIORS|>.comp name) rHPOT")) 

(Is (<|SCENARiO""BEHAV10RS|>TouQ " ("HIGH*!) 

(Is («ISCENARlS BEHAV10RS|».comp name) fMCC COOL")) 

(Is («|SCENARIO"BEHAV10RS|»Touft THIGH^) 

{Is (<«|SCENARl6 BEHAVIORS|>».comp name) fLPFP")) 

{Is (<«|SCENARIO"BEHAV10RS|»>.omefla) 0-OW - )) 
) 

(@HYPO» EXPERT HEURISTIC RULE) 
(©RHS* 

(CreateObject (Randy Hurt NZL CooHngJeak) ({EXPERT RULE SUGGESTIONS] ,\ 

(HEURISTIC SUGGESTIONS!)) 

(Let (Randy Hurt NZL Cooling leak. component name) ("NZL COOL*)} 

(Let (RanoVlHurrNZL"Cooling^leak.suggestedJaulO {"LEAK")} 

(Let (RarKiy~Hurt~NZL~Cooling~leak. suggestion" type) ("specific component")) 
) 
) 

(©RULE* HPOT efficiency low 
@INFCAT»oT 

@COMMENTS="WouW also expect to see HPOT Tout H, and maybe OPB pout H"; 
(@LHS= 

(Is (<ISCENARIO BEHAV10RS|>.comp name) COPOV)) 

(Is (<|SCENARIOlBEHAVIORS|>.posrtton) ("HIGH")) 

(Is («}SCENARIO BEHAVIORS|».comp_name) fHPOP PBP")) 

(Is («|SCENARIO"BEHAVIORS|».omeaa) ("NORMAL")) " 

(Is (<«|SCENARl5 BEHAVlORSl»>.comp_name) ("MCC)) 

(Is («<jSCENARIO"'BEHAVlORS|>».pout) ("NORMAL^) 
) 

(©HYPO* EXPERT HEURISTIC RULE) 
(@RHS= 

(CreateObject (Randy Hurt HPOT efficiency low) ([EXPERT RULE SUGGESTIONS! A 

|HEURISTICSUGGEST10NS|)) " 

"(Let (Rar>dy_Hurt_HPOT_effk;iencyJow.component name) {" HPOT*)) 

(Let (P^ndy^Hurt^HF^T.effciency^lcw.suggestedJauftjn-OW.EFFICIENCYl) 

(Let (Randy~Hurt""HPOT~efftciency"low.suggestiorrtype) ("specific component")) 
) 
) 

(©RULE* match found 2 
@INFCAT=-T5001;~ 

@COMMENTS*"Suggested fault For now, no more than one suggestion Is generated in MADE FIRST SCENARIO!"; 
<@LHS* 

(IsNot (CONTROL OBJECT. matching fault) (NOTKNOWN)) 

(Yes (USE HEURISTIC RULES)) 

(« (LENgTH(<|HEURfSTIC SUGGESTIONS! >)) (1)) 

(. (COMPARE(<|HEURISTlC SUGGESTiONS!>.suggested faultCONTROL OBJECTcurrent fault)) (0)) 

(Reset(SUGGESTION TYPES)) " 

(Tes (SUGGESTfON"TYPES)) 
) 

(©HYPO* MATCH FOUND) 
(@RHS= 

(Strategy (©EXHBWRDsTRUE;)) 
) 
) 

{©RULE* use heuristicsl 
@INFCAT*"-15000; 
(©LHS* 

(is (<!HEURlSTIC_SUGGEST10NSi>.sug9estion type) C*P«rfic_componenn) 

(Name (<1TEMP BEHAVIORl>.comp name) "(CONTROL HEURtSTICS.temporary)) 

(- (CX>MPARE(<lHlURIST1C_SUGGESTIONS|>.c«Tip^ (0)) 

(©HYPO* SUGGESTION TYPES) 
(@RHS- 

Po (1.5) (CONTROL OBJECT.currentjxob)) 

) 

) 

(©RULE* use heuristics 2 
@INFCAT=-15000; 

©COMMENTS*"We are less convinced (prob»1 .0) in this case. TEST: Both comp type and config type must be in the parent classes"; 
(@LHS= 

(Is (<[HEURISTIC SUGGES710NS|>.suggestion type) C«xripcfHirrt and tault .type")) 

(Execute fAtc™NameValue")(@ATOMID*<lTEMP BEHAVIORJ>;@SlBlrTG="^RETURN=CONTROL HEURISTICS. current behavior A 
©NAMES";)) 

(Execute ("GetRelatrves") (@ATOMICU\CONTROL HEURISTICS.current bebavior\;\ 
@STRING*"@CU^SES,@PARENTS,@Ev^YLEVEL.@RETUR^C©^^ 

(Execute (TestMuftiValue") (©ATOMIC *< [HEURISTIC SUGGESTIONSI>. component name;\ 
@STRING.'@SUBSET©TEST-©v(CONTROL HEURISTICS.componerTt classesJA 
@RETURN*CONTROL HEURISTICS. boo) temp,@COMP*STRINGT;)) 
(Yes (CONTROL HEURISTICS. bool temp)) 
) 

(©HYPO- SUGGESTION TYPES) 
(@RHS« 

Po (1.0) (CONTROL OBJECTcurrent jxob)) 



Final Report, July 1993 285 



} 
) 



286 Final Report, July 1993 



Class FUEL_SIDE (File fuel_side) 



\s 



Class FUEL_SIDE (File fuel_side) 



\mcc_cooljstame\="mcc_cool" 

\nzl_cool.name\="nzl cool" 

\fpb.name\="fpb" 

\difpusekname\="diffuser" 

\ftjel_flow_ctrl.name\=*'fuel_flow_ctrl' , 

\mfvJctrl^tame\="nifv_ctrl" 

\ccv_ctrl.name\="ccv "ctrl" 

\LPFfNAME\="LPFT' 

\HPFT.NAME\="HPFT' 

\F101.NAME\=T101" 

\F108.NAME\="F108" 

\F110.NAME\=T110" 

\F109.NAME\=T109" 

\F190.NAME\="F190" 

VFIOT-NAMEVFIO?" 

\HPFP.NAME\="HPFP" 

\LPFP.NAME\= W LPFP" 

\FXJEL TANK.NAME\="FUEL TANK" 

\MKER.NAME\="MIXER" 

\M103.NAME\="M103" 

\FPOV.NAME\=TPOV" 

\MFV.NAME\="MFV" 

\CCV.NAME\="CCV" 

********** 



Final Report, July 1993 PRECEDING PAGE BL,ANK NOT FiLMEU ^ 



Class LOX_SIDE (File k>x_side) 



Class LOX_SIDE (File lox_side) 



\MOV_CTRLJJAME\="MOV_CTRL" 

\OPOV_CTRL.NAME\="OPOV_CTRL" 

VHPOT.NAME\="HPOT" 

\HPOP_PBPJNAME\="HPOP/PBP" 

^LPOTJJAME\='XPOT , 

\OPB.NAME\="OPB" 

\O201AIEDIUM_INPin\="LPOP" 

\O203.NAME\="O203" 

\O204.NAME\="O204" 

\O205.NAME\="O205" 

\O206.NAME\="O206" 

\O190JSrAME\="O190" 

\LPOP.NAME\="LPOP" 

\LOX_TANK.NAME\=1X)X_TANK" 

\M104!nAME\="M104" 

\M101.NAME\="M101" 

\MOV.NAME\="MOV" 

\OPOV.NAME\="OPOV" 

********** 



288 _ Final Report, July 1993 



PBM data.tkb 



A.4 PBM Data Support Files 



PBM data.tkb 



(©VERSION* 020) 

(©PROPERTY* setup ©TYPE-BooJean;) 

(©CLASS* PBM TEMPLATES 
(©PROPERTIES- 

commar>ded_po6ftion 

comp_name * 

h_dlff 

MechPWR 

MR 

omega 

p dm 

pin_OX 

position 

pout 

PV_Product 

q_dot_in 

q dot hB 

T_dffl 

setup 

Tbar 

Tin 

Tin OX 

TinB 

Tout 

Tq 

Tsource 

Vbalance 

Vbar 

VTn 

VTn OX 

Vn5 

Vout 

VoutB 

VoutC 



) 



) 



(©OBJECT* PBM TEMPLATES RLE 
{©CLASSES* 

LOAD_CONTROL 

(©PROPERTIES* 

class name 

file_exists 

file name 

NAME 

retrieve 
) 
) 

(©SLOT. PBM TEMPLATES FlLEdass name 

(@SOURC§S= 

(RunTimeValue rPBM TEMPLATES")) 

) 
) 

(©SLOT* PBM TEMPLATES RLENAME 

(©SOURCES- 

(RunTimevalue (TBM values.nxp'O) 

) 
} 

(©SLOT* PBM TEMPLATES.setup 
(©SOURCES* 

(Execute CAtomNameVaJue") (©ATOM ID* SELF;@STRING*-@ RETURN- CONTROL OBJECT4.temporan/A 
©NAMES";)) 

Po (SL©STRING(CX>NTR0L.OBJECT4.temporan/ ( l (STRLEN{CONTROL_OBJECT4.tempof^ PBM template"))))\ 

(SELF.comp_name)) ~ 

po ~(CONTROL_OBJECT4.temporary) QSELF.comp nameVPBM template_behav»or)) 
(RunTimeValue (TRUE)) 
) 
) 

(©RULE* set_up PBM 

(@LHS- 

(Name (<]PBM_TEM PLATES |>. setup) (<|PBM TEMPLATES|>. setup)) 

) 

(©HYPO* SET UP PBM) 
) 

(©RULE- count PBM matches commanded position 
(@LHS- " 

(Is (\CONTROL OBJECT4.temporaryVcommanded_position) (KNOWN)) 
(Equal(<ICURRENT_BEHAVIOR|>xommanded_position) 0CCNTROL_OBJECT4.temporaryVcommandedj3Osition)) 



Final Report, July 1993 289 



PBM data.tkb 



) 



(@HYPO- COUNT PBM MATCHES) 
(@RHS* 

(Do (<|CURRENT BEHAVIOR|>.temp lrt+ 1) (<|CURRENT BEHAV10R|>.tempJhtJ) 

) 



(©RULE* court PBM matchesjudffi 
(@LHS* 

(Is ^CONTROL OBJECT4.temporarylh_diff) (KNOWN)) 

(Equal(<|CURRENT BEHAVIOR|>.h drff) (\CONTROL_OBJECT4.temporafyVh_dfff)) 

) 

(@HYPO- COUNT PBM_MATCHES) 

(@RHS* 

(Do (-eJCURRENT BEHAVJOR|>.temp kit* 1) (<|CURRENT BEHAV10R1>. temp Jrt)) 

) 



(@RULE= court PBM matehesJfecnPWR 
(@LHS* 

(Is (\CONTROL 08JECT4.tefnporaryVMectiPWR) (KNOWN)) 

(Equal(<|CURRENT_BEHAVIOR|>.M«chPWR) (VCONTROL - OBJECT4.temporaryVMechPWR)) 

(@HYPO- COUNT PBM MATCHES) 
(@RHS- 

(Do (<|CURRENT BEHAVJOR|>.temp int+1) (<|CURREI^J5EHAVIORi>. temp Jrt}) 
) 
) 



(@RULE* court PBM matches MR 
(@LHS« ~ ~ 

(Is (^CONTROL OBJECT4.temporaryVMR) (KNOWN)) 
(Equal(<|CURRENT BEHAVtOR|>.MR) (^CONTROL OBJECT4.temporaryVMR)) 

) 

(@HYPO» COUNT PBM MATCHES) 

(@RHS» 

(Do (<|CURRENT_BEHAVIORi>.temp_lnt + 1) (<|CURRENT_BEHAVIORi>.tempJni)) 



} 



) 



(@RULE= count PBM matches omega 
(@LHS= 

(Is {^CONTROL OBJECT4.temporary\,omega) (KNOWN)) 

(Equal(<|CURR£NT BEHAVIOR^, omega) (\CONTROL OBJ ECT4,temporaryV omega)) 

) " . 

(@HYPO» COUNT PBM MATCHES) 
<@RHS* 

(Do (<|CURRENT_BEHAV)OR|>.temp_int + 1) (<|CURRENT_BEHAVtORI>.temp_lnt)) 



) 



) 



(@RULE^ count PBM matehes_p dtff 
<@LHS« 

(Is (^CONTROL OBJECT4.tsmporaryVp.diff) (KNOWN)) 

(EquaJ(<jCURRENT BEHAV10Ri>.p diff) (\CONTROL_OBJECT4.temporaryVp_diff)) 

) 

<@HYPO* COUNT_PBM MATCHES) 

<@RHS= 

(Do (<(CURRENT BEHAVIOR^. temp int+1) (<|CURRENTJBE!WIORj>. temp Jrt)) 



) 



) 



(@RULE« count PBM matches_pfn 
(@LHS» 

Os (VCONTROt_OBJECT4,temporaryVpin) (KNOWN)) 
(Epual(<lCURRENT BEHAV10R|>.pin) (\CONTROL_OBJECT4.temporaryVpin)) 

) 

(@HYPO* COUNT PBM_MATCHES) 

(@RHS= 

(Do <<|CURRENT_BEHAV10RJ>.tempJnt + 1) (<|CURRENT_BEHAV!Oflj>. temp Jrt)) 



) 



) 



(©RULE* court PBM_mafches_pin OX 
(@LHS= 

(Is (\CONTROL OBJECT4.temporaryVpin.pX) (KNOWN)) 

(Equal (<|CURRENT_BEHAVIOR|>.pinOX) 0CONTROL^OBJECT4.temporafyVpln_OX}) 

(@HYPO- COUNT PBM MATCHES) 
(@RHS* 

(Do (<|CURRENT_BEHAVIOR|>.lempJrt + 1) (<|CURRENTBEHAVIORl>,tempJnt)) 



) 



) 



(@RULE* count PBM matches jwsition 
(@LHS- 

(Is ^CONTROL OSJECT4.temporaryVposHion) (KNOWN)) 
(EquaJ(<|CURRENt - BEHAVIORJ>.position) (CONTROL - 0BJECT4.temporaryVposition)) 

(@HYPO- COUNT PBM_MATCHES) 
(@RH3» 

(Do (<|CURRENT BEHAVIOR|>Jemp int+1) (<|CURRENT_BEHAV10R|>.temp_lrt)) 

) 



290 



Final Report, July 1993 



PBM data.tkb 



1 



(©RULE* count POM matches pout 
(@LHS- " 

(Is (VCONTROL OBJECT4,temporaryVpout) (KNOWN)) 

(Equal(<|CURRENT_BEHAV10Rj>.pout) flCONTROLJ5BJECT4.temporaryVpout)) 

(©HYPO- COUNT PBM MATCHES) 
(@RHS» 

(Do (<)CURRENT BEHAVIOR|>.temp int + 1) (<| CUR RENT BEHAVIORj>.temp WJ) 
) 

} 

(©RULE* count PBM matches PV Product 
(@LHS= " 

(Is (^CONTROL OBJECT4.tamporaryVPV Product) (KNOWN)) 

(Equal (<)CURRENT BEHAVtOR|>.PV Product (\PONTROL OBJECT4.temporaryVPV_Produd)) 

) 

(©HYPO* COUNT PBM MATCHES) 

(@RHS* 

(Do (<|CURRENT BEHAVIOR|>.temp int + 1) (<|CURRENT_BEHAViOR)>.temp Int}) 
) 
) 

(©RULE- count PBM matches q_dotjn 
(@LHS* 

(Is (\CONTROL OBJECT4.temporaryVq_dot in) (KNOWN)) 

(Equa1(<]CURRENT BEHAVIOR) >.q_dot_ln) (fcONTROL 08JECT4.temporaryVq.dot in)) 

) 

(©HYPO- COUNT PBM MATCHES) 

(@RHS* 

(Do (<|CURRENT BEHAVIOR) >. temp kit + 1) (<|CURRENITJ3EHAVI0R|>.temp Int)) 
) 
) 

(@RULE= count PBM matches q dot inB 
(©LHS* " 

(Is (\CONTROL OBJECT4.t»mporaryVeL dot inB) (KNOWN)) 

(Equal(<|CURRENT BEHAVIOR)>.q_dot kfe) (CONTROL 08JECT4.temporaryVq_dot inB)) 

) 

(©HYPO* COUNT PBM MATCHES) 

(@RHS* 

Po (<|CURRENT BEHAVIOR)>.temp_int+1) (<|CURRENT BEHA\/10R|>.temp H)) 
) 
) 

(@RULE= count PBM matches T dfff 
(®LHS* ~ 

(Is (^CONTROL 0BJECT4.temporaryVT ditf) (KNOWN)) 

(Equal(<|CURRENT BEHAVlORJ>.T d'rff) " (\CONTROL OBJECT4.temporary\T drrt)) 

) 

(©HYPO* COUNT PBM MATCHES) 

(@RHS= 

Po (<[CURRENT BEHAVlORl>.tempJnt + 1) (<|CURRENT BEHAVIOR|>.temp k$) 
) 
) 

(©RULE* count PBM matches Tbar 
(@LHS* 

Os CCONTROL OBJECT4.temporary\jDar) (KNOWN)) 

(Equal(<(CURRENT BEHAV(OR|>.Tbar) fliCONTROL OBJECT4.temporaryVTbar)) 

) 

(©HYPO* COUNT PBM MATCHES) 

(@RHS* 

(Do (<|CURRENT BEHAVIOR|>.temp int + 1) (<JCURRENT BEHAV10R|>.temp int)) 



(©RULE* count PBM matches Tin 
<@LHS* 



(is (VCONTROL OBJECT4.temporaryVTin) (KNOWN)) 
(EqualHCURRENT^r - — ~ ' ""- 



BEHAV)OR|>.Tin) (\CONTROL 0BJECT4.tempoTaryVTin)) 
) 

(©HYPCfc COUNT PBM MATCHES) 
(@RHS= 

Po (<|CURRENT BEHAV!OR|>.t»mp knt+ 1) (<]CURRENT BEHAVlOR|>.temp int)) 
) 



(©RULE, count PBM matches Tin OX 
(@LHS* 

(Is (\CONTROL OBJECT4.temporaryVTin OX) (KNOWN)) 

(Equal(<jCURRENT BEHAVIOR|>.Tin OX) "(VCONTROL OBJECT4.terr?waryVTin_OX)) 
) 

(©HYPO* COUNT PBM MATCHES) 
(@RHS« " " 

Po (<|CURRENT BEHAViORi>.temp int ^ 1) (<|CURRENT_BEHAVIOR(>.temp int)) 
) 
) 

(©RULE* count PBM matches TinB 

(©LHS* 

(Is (\CONTROL OBJECT4.temporaryVTmB) (KNOWN)) 

(Equal(<)CURRENT BEHAVlOR|>TmB) ^CONTROL OBJECT4.temporaryVTnB)) 

) 

(©HYPO* COUNT PBM MATCHES) 



Final Report, July 1993 291 



(@RHS« 

(Do (<]CURRENTBEHAV10R|>.tempJnt + 1) 



) 



(<)CURRElsrr_BEI4AVlOR|>.tempjnt)) 



PBM data.tkb 



(©RULE* count PBM matches Tout 
(@LHS« 

(Is (\CONTROL OBJECT4.tempofaryVTout) (KNOWN)) 

(Equal(<|CURRENT BEHAVIORi>.Tout} {^CONTROL OBJECT4.temporanATout)) 



) 



) 

(@HYPOx COUNT PBM MATCHES) 

<@RHS* 

(Do (<|CURRENT BEHAVIOR|>.temp ht + 1) 
) 



(<|CURRENT_BEHAVIOR|>.temp_JrtO) 



(©RULE- count PBM matches Tq 
(@LHS« 

(is (CONTROL OBJECT4.temporaryVTq) (KNOWN)) 

(Equal (<|CURRENT BEHAVIOR|>Tq) (\CONTROL OBJECT4.temporafylTq)) 



) 



) 

(@HYPO= COUNT PBM MATCHES) 

(@RH3= 

(Do (<[CURRENT BEHAVIOR) >. temp Int+1) 
> 



(<|CURRENT_BEHAVIORl>.t»mp_int)) 



(@RULE= count PBM matches_Tsource 
(@LHS* 

(Is (\CONTROL OBJECT4.tempweryVTsource) (KNOWN)) 

(Equal (<|CURRENT BEHAV!OR(>.TsouTce) (\CONTROL OBJECT4.temporaryVTsource)) 



) 



) 

(@HYPO- COUNT PBM MATCHES) 

(@RHSe 

Po (<|CURRENT_BEHAV!ORI>.temp int + 1) 
) 



(<!CURRENT_BEHAV!ORJ>.tBmp_BiQ) 



(@RULE= count PBM matches_Vbalance 
(@LHS- 

(Is 0CONTROL_OBJECT4.temporaryVVbaI»nc«) (KNOWN)) 

(Equal (<|CURRENT BEHAVIOR) >.Vbalance) (\CONTROL OBJECT4.temporaryVVbatance)) 



) 



) 

(@HYPO= COUNT PBM MATCHES) 

<@RHS* 

(Do <<|CURRENT BEHAVIOR) >.temp int+1) 
) 



(<|CURRENT_BEHAVIOR]>.temp_int)) 



(©RULE- count_PBM matches Vbar 
(@LHS= 

(Is (\CONTROL OBJECT4.temp«aryVVbaf) (KNOWN)) 

(Equal (<|CURREN? BEHAVIOR^. Vbar) (CONTROL OBJ ECT4.temporaryV Vbar)) 



) 



) 

(@HYPO* COUNT PBM MATCHES) 

(@RHS= 

(Do (<)CURRENT BEHAVIOR) >. temp ht+1) 

) 



(<[CURRENT_BEHAVJORM 



J«® 



(@RULE= count PBM matches Vin 
(@LHS» 

(Is OCONTROL OBJECT4.temporafyVVin) (KNOWN)) 

(Equal (<)CURRENt BEHAVlORj>.V]n) 0CONTROL_OBJECT4.temporanAVVi)) 



) 



(@HYPO- COUNT PBM MATCHES) 
(@RHS* 

(Do (<jCURRENT BEHAViOR|>.temp int+1) 
) 



(<|CURRENT_BEHAVIOR|>.temp_int)) 



(©RULE* count PBM matches Vin OX 
(@LHS= 

(Is <KX3NTTOL_OBJECT4.temporafyVVin OX) (KNOWN)) 

(Equal (<1CURRENT BEHAVIOR)>.V*n_OX) ~(\f»NTROL_OBJECT4.tempofaryVVin_OX)) 



) 



<@HYPO« COUNT PBM MATCHES) 
(@RHS= 

(Do (<|CURRENT_BEHAVIOR|>.temp_int + 1) 

) 



(<|CURRENT_BEHAVIOR|>.temp_inQ) 



(©RULE- count PBM matches VmB 
<@LHS» " 

(Is (\CONTROL OBJECT4.tempofaryVVinB) (KNOWN)) 

(Equal(<)CURRENT BEHAVIOR]>.VinB) ^CONTROL OBJECT4.tempor«yVVinB)) 



) 

(@HYPO= COUNT PBM MATCHES) 

(@RHS« 

(Do (<)CURRENT BEHAVlOR|>.temp int+1) 
) 



(<iCURRENT_BEHAV10R|>.t»mp Jn§) 



292 



Final Report, July 1993 



PBM data.tkb 



(@RULE= count PBM matches Vout 
{©LHS- 

(ls OCONTROL , OBJECT4.temporaryVVout) (KNOWN)) 

(Equal(<|CURRENTBEHAVIOR|>.Vout) (\OONTROL_OBJECT4.temporafyVVoot)) 

<@HYPO COUNT PBM MATCHES) 
(@RHS* 

(Do (<|CURREWT_BEHAVIOR|>.tempJnt + 1) (<|CURRENT_BEHAVtOR|>.temp>l3) 

) 

(@RULE= count PBM matches VoutB 
(@LHS* " 

(Is OCONTROL . OBJECT4.temporan/VVoufi) (KNOWN)) 
{Equaf(<lCURRENT_BEHAVIOR|>.Vou!B) 0CONTROL_OBJECT4.temporanAVoutB)) 

(@HYPO* COUNT PBM MATCHES) 
(@RHS= " 

{Do (<|CURRENT_BEHAVIOR[>.temp - H+ 1) <<|CURRENT_BEHAVIOR|>.t»mpJnt)) 

) 

(@RULE« count PBM matches VoutC 
(@IHS= " " 

(Is (\CONTROL OBJ ECT4.temporary\, VoutC) (KNOWN)) 
(Equal(<[CURRENT_BEHAV10R|>.VoutC) <\CONTROLJ)BJECT4.t»mp«aryWoutC)) 

(@HYPO= COUNT PBM MATCHES) 
(@RHS* 

(Do (<lCURRENT_BEHAVtOR|>.temp.int + 1) (<|CURRENT_BEHAVIORI>,ternpJnt)) 

) 



Final Report, July 1993 293 



PBM Parameters 



PBM Parameters 



[MOVp dffq 

[0201 p dffi| 
[LPOTp diffj " 



[MFVp_dfff] 



2 RMEP ENGINE MIXTURE RATIO [MCC MR] 

4 P1FP1 LPFP INLET PRESSURE [LPFP pin] 

5 P10P1 LPOP INLET PRESSURE [LPOPpIn] 

6 T1FP1 LPFP TEMPERATURE [LPFP Tin] 

7 T10P1 LPOP INLET TEMPERATURE [LPOP Tin] 
10 PCCOM CHAFER PRESSURE COMMAND 
227 DP21 MOV INLET P RESS URE DROP 
243 DPOCO LPOP DISCHARGE DUCT PRESSURE DROP 
246 DPOT1 LPOT PRESSURE DROP 

254 DP{1) HPFP DISCHARGE DUCT PRESSURE DROP 

255 DP{2) PBP INLET DUCT PRESSURE DROP 
256DP(3) MFV PRESSURE DROP 
257 DP 4) PBP DISCHARGE DUCT PRESSURE DROP 
25S DP 5) NOZZLE JACKET DISCHARGE MANIFOLD & MIXER PRESSURE DROP 
259 DP(6) MOV PRESSURE DROP [MOVp_dHfl 
262DP(9) CCV INLET DUCT PRESSURE DROP 

263 DP(10) HPOP DISCHARGE DUCT PRESSURE DROP - SECTION 1 

265 DP(12) OPB OXIDIZER INLET DUCT PRESSURE DROP 

266 DP(13) OPB FUEL INLET MANIFOLD PRESSURE DROP 

267 DP(14) OPOV PRESSURE DROP [OPOVp_dttQ 

270 DP(17) FPB FUEL MANIFOLD PRESSURE DROP 

271 DP(18) FPOV PRESSURE DROP [FPOVpjflff] 

274 DP(21) HPOT INLET DUCT PRESSURE DROP 

275 DP(22) LPFT DISCHARGE DUCT PRESSURE DROP 

277 DP(24) HPFT INLET DUCT PRESSURE DROP 

278 DP(2S) LPFT INLET DUCT PRESSURE DROP 

279 DP{26) HPFT COOLING CIRCUIT PRESSURE DROP 

281 DP{28) MCC COOLING JACKET DISCHARGE MANIFOLD PRESSURE DROP 

282 DP(29) LPOT INLET DUCT PRESSURE DROP - SECTION 2 

283 DP{30) MFV DISCHARGE DUCT PRESSURE DROP 

285 DP(32) FPB FUEL INLET DUCT PRESSURE DROP [F1 1 p_diff] 

286 DP(33) MCC COOLING JACKET INLET DUCT PRESSURE DROP 
288 DP{35) NOZZLE COOLING JACKET INLET DUCT PRESSURE DROP 

290 DP{37) CCV PRESSURE DROP [CCV p_dtfT) 

291 DP(38) CCV DISCHARGE DUCT PRESSURE DROP 

292 DP(39) FPB FUEL INLET DUCT PRESSURE DROP 
294 DP(41) OPB FUEL INLET DUCT PRESSURE DROP 

305 DP(52) NOZZLE COOLING JACKET DISCHARGE DUCT PRESSSURE DROP 

313 HINJGF HGM HOT GAS ENTHALPY -FUEL SIDE 

31 4 HINJGO HGM HOT GAS ENTHALPY - OXIDIZER SIDE 

31 5 DTBYO OPB COOLING CIRCUIT TEMPERATURE RISE 

316 DTMFV MFV TEMPERATURE RISE 

317 DTFP1 LPFP TEMPERATURE RISE 

318 DTFP2 HPFP TEMPERATURE RISE 

31 9 DTFT1 LPFT TEMPERATURE RISE 

320 DPS9 CCV DISCHARGE DUCT & MIXER PRESSURE DROP 

322 DTFT2 HPFT TEMPERATURE DROP 

323 DTOT2 HPOT TEMPERATURE DROP 

324 H10T1 LPOT INLET ENTHALPY 

325 H20T1 LPOT DISCHARGE ENTHALPY 
328 H20T2 HPOT DISCHARGE ENTHALPY 

334 DTJ MCC NOZZLE JACKET TEMPERATURE RISE [NZL OOOLT.difg 

335DTJ2 MCC JACKET TEMPERATURE RISE [MCCCOOl Tjjrff] 

336 DTOP2 HPOP TEMPERATURE RISE 

337 DTOP3 TOP TEMPERATURE RISE 

359 ENFT1 LPFT SPEED 

360 ENFT2 HPFT SPEED 

361 FECOM ENGINE THRUST (COMMANDED) 

362 FSL ENGINE THRUST (SEA LEVEL) 

364 ENOT1 LPOT SPEED 

365 ENOT2 HPOT SPEED 

369 H1CCV CCV INLET ENTHALPY 

370 H38 CCV DISCHARGE ENTHALPY 
MIXER HOT HYDROGEN 'INLET ENTHALPY 

NOZZLE COOLING JACKET INLET ENTHALPY 
NOZZLE COOLING JACKET DISCHARGE ENTHALPY 
MCC COOLING JACKET INLET ENTHALPY 

MCC COOLING JACKET ENTHALPY AT BOUNDARY LAYER ATTACH POINT 
LPFP HEAD RISE [LPFP p_dfflj 

_. . HPFP HEAD RISE [HPFP p_diffj 

380 H2FT2M HPFT DISCHARGE ENTHALPY (AFTER MIX) 

381 HINJ MAIN INJECTOR HOT GAS ORIRCE INLET ENTHALPY 

382 HINPB1 NOZZLE COOL JACKET DISCHARGE ENTHALPY (AFTER MIX) 

383 HINPB2 PREBURNER INLET FUEL EFFECTIVE ENTHALPY 

384 HOP1 LPOP HEAP RISE [LPOP p drff] 

385 HOP2 HPOP HEAD RISE [HPOP_PBP p_dffi] 

386 HOP3 PBP HEAD RISE 

387 HPFP1 LPFP POWER 

388 HPFP2 HPFP POWER 

389 HPFT1 LPFT POWER 
HPFT POWER 
LPOP POWER 
HPOP POWER 
PBP POWER 
LPOT POWER 
HPOT POWER 

396 HCT HPFT COOLING CIRCUIT DISCHARGE ENTHALPY 

397 H20T2M HPOT DISCHARGE ENTHALPY (AFTER MJ)Q 

398 H1MFV MFV INLET ENTHALPY 

399 HTBYO OPB COOLING CIRCUIT DISCHARGE ENTHALPY 

400 H1FP1 LPFP INLET ENTHALPY 

401 H1FP2 HPFPJNLET ENTHALPY 

402 HtFfl LPFT INLET ENTHALPY 

403 H1FT2 HPFT INLET ENTHALPY 

404 H10T2 HPOT INLET ENTHALPY 



371 H2J1 

372 H1 

373 H2 

374 H3 

375 H4 

378 HFP1 

379 HFP2 



390 HPFT2 

391 HPOP1 

392 HPOP2 

393 HPOP3 

394 HPOT1 

395 HPOT2 



[LPFP MechPWR] 
[HPFP MechPWR] 
[LPFT MechPWR] 
[HPFT MechPWR] 
[LPOP MechPWR] 
rHPOP_PBP MechPWR] 

[LPOT MechPWR] 
[SPOT MechPWR] 



294 



Final Report, July 1993 



t y 



405 H2FP1 LPFP DISCHARGE ENTHALPY 

406 H2FP2 HPFP DISCHARGE ENTHALPY 

408 H2FT1 LPFT DISCHARGE ENTHALPY 

409 H2FT2 HPFT DISCHARGE ENTHALPY 

412 H2GP3 PBP DISCHARGE ENTHALPY 

413 H20F2 HPOP DISCHARGE ENTHALPY 

41 4 PEXC HGM DISCHARGE PRESSURE - FUEL SIDE 

415 PEXCO HGM DISCHARGE PRESSURE - OXIDIZER SIDE 
417 PEXTCJ PREBURNER FUEL SUPPLY DUCT INLET PRESSURE (MIXER DISCHARGE) 

420 PFPB FPB CHAMBER PRESSURE [FPB pout HPFT pinj 

421 H20P1 LPOP DISCHARGE ENTHALPY 

422 H10P1 LPOP INLET ENTHALPY 

423 H10P2 HPOP INLET ENTHALPY 

424 H10P3 PBP INLET ENTHALPY 
433 PINFM HGM INLET PRESSURE -FUEL SIDE [HGMpinJ 

442 POMC HGM COOLING CIRCUIT INLET PRESSURE - OXIDIZER SIDE 

443 POPB OPB CHAMBER PRESSURE [OPB pout] 
458 P1MFV MFVINLETTOTAL PRESSURE [MFVpm] 

460 P1FP2 HPFP INLET PRESSURE JF101 poul^ 

461 P1FT1 LPFT INLET PRESSURE [LPFT pin] 

462 P1FT2 HPFT INLET PRESSURE [HPFT pin] 

463 P10P3 PBP INLET PRESSURE 

464 P10P2 HPOP INLET PRESSURE [HPOP PBP pin 0201 pout] 

465 P1 OT1 LPOT INLET PRESSURE [LPOT pin M1 04 pin] 

466 P1CT2 HPOT INLET PRESSURE [HPOTpin] 
469 P1 HPOP DISCHARGE DUCT DISCHARGE PRESSURE 

476 P1 2 OPOV INLET PRESSURE [OPCV pin M1 01 pin] 

477 P1 3 OPB FUEL INJECTOR INLET PRESSURE 

478 P1 4 OPB OXIDIZER INJECTOR INLET PRESSURE 

481 P17 FPB FUEL INJECTOR INLET PRESSURE 

482 P18 FPB OXIDIZER INJECTOR INLET PRESSURE 

485 P2FP1 LPFP DISCHARGE PRESSURE 

486 P2FP2 HPFP DISCHARGE PRESSURE 

487 P2CP1 LPOP DISCHARGE PRESSURE 

488 P2CP2 HPOP DISCHARGE PRESSURE 

489 P20P3 PBP DISCHARGE PRESSURE 
493 P210 MOV INLET PRESSURE 
496 P3P MFV DISCHARGE PRESSURE 

501 P2TFT1 LPFT DISCHARGE PRESSURE 

502 P2TFT2 HPFT DISCHARGE PRESSURE 

504 P2TOT2 HPOT DISCHARGE PRESSURE 

505 P20 CCV INLET PRESSURE 
515 P2MFV MFV DISCHARGE PRESSURE 

519 P33 MCCCOOUNG JACKET INLET PRESSURE 

520 P34 OPOV DISCHARGE PRESSURE 

521 P35 NOZZLE COOLING JACKET INLET PRESSURE 

522 P36 FPOV DISCHARGE PRESSURE 

523 P37 CCV DISCHARGE PRESSURE 

525 F39 FPB FUEL MANIFOLD INLET PRESSURE 

526 P4 PBP DISCHARGE DUCT DISCHARGE PRESSURE 

528 P41 OPB FUEL MANIFOLD INLET PRESSURE 

529 P42 OPB IGNITER OXIDIZER INLET PRESSURE 
535 P6 MOV DISCHARGE PRESSURE 
537 P7 MFV DISCHARGE DUCT DISCHARGE PRESSURE 
539 P9 PREBURNER FUEL SUPPLY DUCT DISCHARGE PRESSURE (FINAL) 

558 RMG FPB MIXTURE RATIO [FPB MR] 

559 RMGO OPB MIXTURE RATIO [OPB MR] 
583 TCT HPFT COOLING DISCHARGE TEMPERATURE 

585TEXTCJ MIXER DISCHARGE TEMPERATURE [MIXER Tout F1 07 Tin] 

586T2CCV CCV DISCHARGE TEMPERATURE [CCV Tout MIXER Tin] 

594 TJ1 NOZZLE COOLING JACKET DISCHARGE TEMPERATURE [NZL COOL Tout MIXER TViB] 
596 T10P2 HPOP INLET TEMPERATURE [HPOP PBP~TinO201 Tout] 

604 TORFT1 LPFTTORQUE 

605 TORFT2 HPFT TORQUE 

606TOROT1 LPOTTORQUE [LPOT Tq LPOP Tq] 

607 TOROT2 HPOT TORQUE 

608 TPB FPB TEMPERATURE [FPB Tout HPFT Tm] 
609TPBO OPB TEMPERATURE [OPB Tout HPOT Tin] 
61 3 TTBYO OPB COOLING CIRCUIT DISCHARGE TEMPERATURE 
615T1FP2 HPFPINLETTEMPERATURE [HPFP Tin F1 01 Tou^ 

LPFT INLET TEMPERATURE [LPFT Tm] 



[MIXER pin F1 07 pin] 



[OPB ptn_OX] 

[FPB pin OX] 

[LPFP pout F1 01 pin] 

[HPFP pout] 

[LPOP pout 0201 pin] 
[HPOP.PBP pout] 

[MOV pin O204 pouq 
[MFV pout DIFFUSERpin] 
[LPFT pout] 

[HPFT pout] 

[HPOT pout] 
[CCV pin] 

[MCC COOL pin] 
[OPOV pout]" 

[NZL COOL pin] 
[FPOV pout]" 
[CCV pout] 

[FPB pin F1 10 pout] 

[OPB pin Ft 08 pout] 
[MOV pout MCC pin.OX) 



616 T1FT1 
621 T2FP1 

626 T2FP2 

627 T2FT1 

628 T2FT2 



[LPFP Tout F1 01 Tm] 
[HPFP Tout MFV Tm] 
[LPFT Tout] 
[HPFT Tout HGM Tm] 

[LPOP Tout O20lTin] 
[HPOP_PBP Tout M104 Tm] 

[LPOT Tout] 
[HPOTToutHGMTinB] 
[MFV Tout DIFFUSERTm] 
[FPOV Tout FPB Tm OX] 

[OPOV Tout OPB Tm OX] 



LPFP DISCHARGE TEMPERATURE 
HPFP DISCHARGE TEMPERATURE 
LPFT DISCHARGE TEMPERATURE 
HPFT DISCHARGE TEMPERATURE 
629 T2FT2A HPFT DISCHARGE TEMPERATURE (MIXED) 

633 T20P1 LPOP DISCHARGE TEMPERATURE 

634 T20P2 HPOP DISCHARGE TEMPERATURE 
635T20P3 PBP DISCHARGE TEMPERATURE 
636T20T1 LPOT DISCHARGE TEMPERATURE 
637 T20T2 HPOT DISCHARGE TEMPERATURE 
639 T2MFV MFV DISCHARGE TEMPERATURE 
649T2FCV FPOV DISCHARGE TEMPERATURE 
650T2OCV OPOV DISCHARGE TEMPERATURE 
661 WFE ENGINE FUEL ROWRATE 
663WFJ2 MCC JACKET FLOWRATE 
665WMFV MFV FLOWRATE 

674 WFPBIF FPB IGNITER FUEL FLOWRATE 

675 WFPBIO FPB IGNITER OXIDIZER ROWRATE 

677 WFPBF FPB FUEL FLOWRATE 

678 H1J2 MCC COOLANT INLET ENTHALPY 

679 WFPBO FPB OXIDIZER FLOWRATE 

680 WFP1 LPFP FLOWRATE 

681 WFP2 HPFP FLOWRATE 

682 H2J2 MCC COOLANT DISCHARGE ENTHALPY 
684 WFTEX HGM MANIFOLD ROWRATE - FUEL SIDE [HGM Vin HPFT VoufJ 

686 WFT1 LPFT ROWRATE [LPFT Vm F1 09 Vout] 

687 WFT2 HPFT ROWRATE [HPFT Vm FPB Vout] 



[MFV Vbar DIFFUSER Vin] 



[FPBVmFHOVout] 

[FPBVm OX FPOV Vout] 
[LPFP Vbar] ~ 
[HPFP Vbar] 



Final Report, July 1993 



295 



689 WINMC LPFTDISCHARGE DUCT FLOWRATE 

692 WMOV MOVFLOWRATE P^V Vbar 0264 Vout MCCVin_OX] 

701 WOE ENGINE OXIDIZER FLOWRATE 

707 WOPBF OPB FUEL FLOWRATE [OPB Vin F1 08 Vout] 

700 WOPBR OPB FUEL INLET DUCT FLOWRATE 

709 WOPBO OPS OXIDIZER ROWRATE (ORB Vin OX OPOV Vout] 

711 WOP1 LPOP FLOWRATE [LPOP Vbar 02di Vm] 

712 WOPlEX LPOP DISCHARGE DUCT FLOWRATE (O201 Vbar] 

71 3 WOP2 HPOP INLET FLOWRATE (NOT INCLUDING WBYOF3 OR WOTEB)[HPOP PBP Vbar O201 Vout M1 04 Vin] 

714 W0P2EX HPOP DISCHARGE FLOWRATE 

71 5 WOPS PBP FLOWRATE (INCLUDES BYPASS ROW & BEARING COOLANT FLOW) 
720 WOT1 LPOT FLOWRATE [LPOT Vbar O203 Vout] 
723W10P3 PBP INLETDUCT FLOWRATE- SECTION 1 _. 

724 W21 HPOP DISCHARGE DUCT FLOWRATE - SECTION 2 

725 WOP3P PBP FLOWRATE (INCLUDES BYPASS FLOW & BEARING COOLANT FLOW) 
729 W0T2 HPOT FLOWRATE [HPOT Vin OPB Vout] 

731 WPBO PREBURNER COQDI^R FLOWRATE 

739 WTCUBY NOZZLE JACKET BYPASS FLOWRATE 

740 WTCJ1 NOZZLE JACKET FLOWRATE [NZL COOL Vbar} 
747 W22 LPFT DISCHARGE DUCT FLOWRATE 

766 W1 3 OPB FUa INLET MANIFOLD FLOWRATE 

767 W1 7 FPB FUEL INLET MANIFOLD FLOWRATE 

768W514 OPOV FLOWRATE [OPOV Vbar M101 Vout] 

769 W518 FPOVFLOWRATE [FPOV Vbar O206 VouQ 

772 W30 MFV DISCHARGE DUCT FLOWRATE 

795 POSMFV MFV POSITION 

796 POSMOV MOV POSITION 

797 POSCCV CCV POSITION 
796 POSOPB OPOV POSITION 
799 POSFPB FPOVPOSFTION 

807 WMIX2 MIXER FLOWRATE (NOZZLE) [MIXER VinB NZL.COOL Vout] 

808 WPBFU MIXER DISCHARGE FLOWRATE [MIXER Vout F1 07 Vin] 
813 WOPBHG OPB HOT GAS FLOWRATE [OPB Vout HPOT Vm] 

908 D{8) LPFT SPEED 

909 D(9) HPFT SPEED 

910 0(10) HPFT DISCHARGE TEMPERATURE AVERAGE 

911 0(11) FPB CHAMBER PRESSURE 

913 D{13) LPOT SPEED 

914 D(14) HPOP DISCHARGE TEMPERATURE 

916 D(16) HPOTSPEED 

917 0(17) HPOT DISCHARGE TEMPERATURE AVERAGE 
91 fi D(18) OPB CHAMBER PRESSURE 

920 D(20) LPFT INLET TEMPERATURE 

969 0(69) HPFP DISCHARGE TEMPERATURE 

990 D(90) NOZZLE COOLING JACKET DISCHARGE TEMPERATURE 

991 D(91) HPOP INLET TEMPERATURE 

992 D(92) PBP DISCHARGE TEMPERATURE 

993 D(93) MIXER DISCHARGE TEMERATURE 

994 0(94) LPFP DISCHARGE TEMPERATURE 

995 D(95) LPFT DISCHARGE TEMPERATURE 

997 D(97) OPOV DISCHARGE TEMPERATURE 

998 D(98) FPOV DISCHARGE TEMPERATURE 

1104WFT1C MCC COOLANT ROWRATE (MCCCOOL Vbar] 



2§>6 Final Report, July 1993 



displaydb.c 



displaydb.c 



#define SUN r this may also work on HP9k */ 

#inckjde <stdio.h> 
#inchjde "dbaccess.h" 

idefine MAXINDEX 1349 
#define FSIZE sizeofffloat) 

RLE*fp,*fp2; 

char *datafilenamc, *datafilename2; 

int slice, index; 



int printjtem (index, value) 
int index; 
float value; 

i 
char mys1r[80]; 
switch (index) { 
r case 933: 

strcpy(mystr, (char)value); 

printfCTest number: %%", myslr); 

break;*/ 
case 934: 

printfCTest date: %6.0fW, value); 

break; 
case 935: 

prtntffTest duration %3.0fvr, value); 

break; 
case 936: 

printffSlice duration %3.0fVi", value); 

break; 
case 937: 

printffSlice start time %3.01\n", value); 

break; 



default 
printfflndex %d Value %YT , index, value); 
break; 



) 



int prirrtjtem2(index, valuel , values) 

int index, 

float valuel , value2; 
( 
char mystr[80]; 
switch (index) { 
r case 933: 

strcpy(mystr, (char)value1); 

printffTest number: %s", mystr); 

break; 7 
case 934: 

printfCTest date: %6.0f\n", valuel); 

break; 
case 935: 

printfCTest duration %3.0nn", valuel); 

break; 
case 936: 

printffSlice duration %3.0ftn'\ valuel); 

break; 
case 937: 

printffSlice start time %3.0rVi", vaJuel); 

break; 



default 

if ((valuel > 0.0000001 ))|(value1 < -0.0000001)) { 
printfflndex %5d Valuel %10f Value2%l0f Dffi%12f %% %5.2fK 

index, valuel , vaJue2, value2-value1 , (value2-value1 )/vakje1 *1 00.0); 
} 

e!se{ 
printfflndex %5d Valuel %10f Vaiue2%l0f Diff%l2fYT, 
index, valuel , value2, vaJue1-value2); 
} 
break; 

) 



) 



main(argc,argv) 
int argc; 
char *argvQ; 

{ 

float a array [1 350); 
float a_array2[1 350]; 
int i; 

if (argc <* 3) {printff Usage: displaydb <datajile_name> <slice #> <index>V0; 

printff or: displaydb <data_file.1>"<data_1ile 2><slice #><index>VO; 

printff If index < 0: lists all values in siTceVfy 

printff If 2 file names are given (second form): Values from both are printedYO; 



Final Report, July 1993 



297 



displaydb.c 



printff and their absolute difference and percent change are shown .Vi"); 
«dt(1);} 



datafilenarne - argv[1]; 

rf (argc ■■ 5) { r two files given ■/ 

datafilename2 - argv[2J; 

si fee * atoi{argv[3]); 

index * atoi(argv[4J); 

) 
eise 

{ 
slice = atoi(argv[2]); 
b atoi(argv[3]); 



:"a1613"; 



rf(argc«5){ 
prlntfCyM»»**»»"«*«««**======= ■ = 

printffSlk* %5d Rle1 %11s Rle2%11s %17s %6sVi", slice, datafilename, datafUenameS, "Dtff 2-1 ", "%"); 






printfC- 

> 
erse{ 

printffRle %$ Slice %d Index %dVi", datafilenarne, slice, index); 

printfC \PVO; 

} 



-V1V1"); 



if ({ip = fopen(datafilename, V)} == NULL) { 

^printffstderr, "%s: Cannot open %sVi\ *argv, datafiJename); 

exit(1); 
> 



tf(argc==5){ 

if {(1p2 = fopen(datafilename2 ( "r")} == NULL) { 
tprintf{stderr, "%s; Cannot open %sVi", *argv, datafilenameS); 
exrt(1); 
} 
) 



if (tseekft), (slice-1)*FSIZETMAXJNDEX +1), SEEK_SET) != 0) { 
printf{" Error seeking time slice in file %syv, datafilenarne); 
exit(1); 

} 

rf (argc -- 5) { 
if {tseekft>2, (slice- 1)*FSiZE*(MAXINDEX +1), SEEK_SET) I- 0) { 
pflntffError seeking time slice In file %sVi", datafilename2); 

> 
> 



fread(a array, FSIZE, MAXINDEX + 1, 1p); 
for (i * 5; i <* MAXINDEX; i++) { 

a arrayp] = xflt(a array+Q; 
) " 



if (argc ■■ 5) { 

*read(a array2, FSIZE, MAXINDEX + 1, fp2); 

for (i = 0; i <= MAXINDEX; i++) { 
a array2fji = xftt(a array 2+ i); 

} * 
} 



if (argc »* 5) { 
if (index < 0) { 

for (i = 0; i <* MAXINDEX; i++) { 
if ((a arrayp] > 0.0000001)||{a_array[i] < -0.0000001)) { 
prTntff%10d : %17f%l81%ief %9,2fy)*, 

i+1 , a_arrayp], a_array2{t], a_array2[i]-a__array[ij» 
(a_arrBy2{i]-a_aSaypI)/a_an-ayfi}*1 00.0)' 

eisef 

printfr%10d : %17f %18f %18fW\ 

M, a arrayp], a array2[i], a array pj-a array2p]); 
} 
} 
) 

else{ 

print item2(index, a array[index-1], a array2[index-1]); 
} 
} 
else 

{ 

if (index <0){ 
for (i * 0; i <« MAXINDEX; i++) { 

printff%i : %fln", i+1 , a arrayfj]); 
} 
} 

else{ 
prirrtjternflndex, a_arTayTmdex-1]}; 



298 



Final Report, July 1993 



displaydb.c 



> 

fclose(1p); 

if (argc ■» 5) { 
tdose{1p2); 
} 

printfOVi"); 

} 



Final Report, July 1993 299 



make_PBM_params.c 
make_PBM_params.c 



^include <stdio.h> 



char *get_qual_vaiue(FlLE *devJHe, int loc); 

Int wrfteJtoJile"{RLE ^r«xp_file ( ^char "comp, char -param, char "q_val); 

const char *kw"LOW"; 
const char * normal="NORMAL"; 
const char * high*"HIGH"; 

int write to file{RLE *nexp file, char *comp, char "param, char *q_vaf) 

{ 

/■prtntfffttts PBM_tBmplate.%s\\-\"^s\ , Vi" ) comp i param ( q_val); */ 

fcrihtf(rwxp3lO\%sJ^MJerry>late.%s\VTOsW 

char «get qual vatue(RLE *dev file, int loc) 

{ 
extern const char * low; 
extern const char * normal; 
extern const char * high; 
•ntlocZ; 

char devjine(256], Result; 
float percent^change, valid, limrU2.5; 

do{ 

fgets(dev line, 256, dev file); 

sscanffdev line,' %d ", &oc2); 
> 

while (ioc2 1* toe); 

sscanf{dev Jine," %*d : %f %*! %*f %f ', &valid, &percent_change); 
r printff%f? w , percent change); */ 
if (valid != 0-0) { 

if (percent_change > limit) result * high; 

else H (percent_ch*nge < (0 - limit)) result * low; 

else result * normal; 

return (result);} 
else return (NULL); 
) 

int main (int argecnar *"argv) 

{ 

r const char * low-"LOW; 
const char • normaUTSIORMAL"; 
const char • high«"HIGH"; */ 

int i. location; 

RLE 'deviation file, -parameter file, "nexpert file; 
char devjine[2563, par_line{2S6i comp[80J, param[80]; 
char bracket, "q^valueT 

if (! (devtat»on_file = kx>enf reMjiumeric_deviaUons", "r"))) 

printf{"Carmot open input file PBM numeric deviations !VV); 
exit(1); 

}; 

If (I (parameter_fite * fopenfPBM_parameters", V})) 

printffCannot open Input file PBM_parameters!VO; 

fclose(deviation file); 

exit(1); 

}; 

if (! (nexpert file * fopenCPBM_values.nxp", %/))) 
{ 
printf("Cannot open output file PBM_v alues. nxpSVO; 
fclose{deviationJtle); ~ 

fctose(parameter file); 
exit{1); 

}; 

for (i=1 ; i <m 5; i+t-) fgets{devjine, 256, deviation_file); 

while {¥gets{par line, 256, parameter_file)) 

{ 
H (bracket = (char *}strchr{parJine,T)) 

( 

sscanf(par_line,"%d ", allocation); 
qjralue = get qua! valuefdeviation file, location); 
if{q_vaJue}{~ 
bracket**; 

sscanf(bfacket,''%s'\comp); 
bracket +* strfen{comp) + 1 ; 
sscanf (bracket "%s",param) ; 
bracket +* strlen (param); 
while r(tracket-1)l=T){ 

r prtntfClOCATlON %d %s %s\n", location.comp.param); */ 
wr ite_to^fne{nexpert_file, comp,param,q_vaiue) ; 
bracket**; 

sscanf(bracket' , %s",comp); 
bracket *■ strton{comp) + 1 ; 



300 Final Report, July 1993 



make_PBM_params.c 



s*canf(bracket,' , %$" 1 param); 
bracket +« sfrlen(param); 

} 

"(param+strien(param)-1) * ~\0'; 

r prtotfO-OCATION *d %s %sVi", location, comp.param}; */ 
write to file(nexpeft file,comp,param,q_value); 
} 



1p r intf (nejq»ert_fite, "■ 



fck»e(deviat'on_flte); 
1ck)se(pa/ameter_fiie) ; 
fck»e(nejq?ert_fiie); 

exit(O); 
} 



Final Report, July 1993 301 



Qualitative Parmeter Values 



A.5 MCC Leak Example Case Data 



A.5.1MCC Leak Example: Qualitative values of measured parameters 

Qualitative Parmeter Values 

\MCC_PC_REFERENCE.QUALITATIVE_VALUE\="NORMAL" 

NMCC.PC.QUALITAriVE^yULUEV^'NORMAL" 

\LPFP_DS_PR.QUALrTAIlVE_VALUE\="LOW" 

\LPFP_DSjrNlP.QUALITAriVE_VALUE\="NORMAL" 

\HPFP_DS_PR.QUALITAnVE VALUE\="NORMAL" 

\MCCjCIJSrTJDS_PR.QUALlTAriVE_VALUE\="LOW" 

\MCCjajNrr_DSjn^.QUALrrATIVE_VALUE\="LOW'' 

\FPB_PC.QUALITAnVE_VALUE\="NORMAL" 

\HPFP_CLNT_LNR_PR.QUALITAnVE VALUE\="NORMAL" 

\MCC_FUELJNJECTOR_PR.QUALm5TVE_VALUE\="NORMAL" 

VLPOP_DS_PR.QUALITAnVE VALUE\="NORMAL" 

\HPOP DS_PR.QU^rrAnVE~VALUE\="NORMAL" 

\LOX_DOME_TMP.QUALITAnVE VALUE\="NORMAL" 

\PBP_DS_PR.QUALrTAnVE_VALUE\="NORMAL" 

\PBP_DS_TMP.QUALITAnVE_VALUE\="NORMAL" 

\HPOP DS_TMP.QUALrrATIVE_VALUE\="NORMAL" 

VENG FUEL_INLETjn^.QUALITAITv^_VALUE\="NORMAL" 

\^G^FUEL_INI^T>R.QUALrrAnVE_VALUEy="NORMAL" 

\ENG_OX_INLET_TMP.QUALITAnVE VALUE\="NORMAL" 

\ENG_OX_INLET_PR.QUALITAnVE_VALUE\="NORMAL" 

\FUEL_FLOW.QUALn , ATIVE_VALUE\="NORMAL" 

\HPFP_DSJ[MP.QUALITA^IVE_VALLFE\="NORMAL' , 

\HPFP_BAL_CAV_PR.QUALITATIVE VALUE\="NORMAL" 

\HPFP COOLANT LINER TMP.QUALITATIVE VALUE\="NORMAL" 

\HP1T_DRAIN_PR.QUALITATIVE_VALUE\= , "N6RMAL" 

\HPFP_DRAIN_TMP.QUALITAnVE VALUE\="NORMAL" 

\LPFT_INLET_PR.QUAL^^AT^VE_VALUE\= ,, LOW ,, 

\FUEL_PRESSURAm , _INTERFACE_PR.QUALITATIVE_VALUE\= ,, NORMAL'' 

\FTJEL_PRESSURANT_INTERFACE_TMP.QUALrrATIVE VALUE\="NORMAL" 

\OPB PCQUALrrATTVE VALUE\="NORMAL M 

\FPOV_POSrnON.QUALTrAnVE_VALUE\="NORMAL" 

\MCC_LOX_INJECfORJIMP.QUALrTAtlVE_VALUE\="NOW^^ 

\MCC_LOX_INJECTOR_PR.0UALrrATiVE3ALUE\= ,s N0RMAL ( ' 

\POGO_PRE_CHARGE_PR.QUALrrAnVE_VALUE\="NORMAL" 

\HEAT_EXCHANGER DS_PR.QUALITAnVE_VALUE\="NORMAL" 

\HEAT_EXCHANGERIINTERFACE_PR.QUAL1TATIVE_VALUE\= , 'N0RMAL , ' 

\HEAT_EXCHANGER_INTERFACE_TMP. % QUALITATrVE_VALUE\=''NORMAL'' 

\OPOV_POSITION.QUALITATrVE_VALUE\= "NORMAL" 

\HPOP_SPEED.QUALITATIVE_VALUE\=''NORMAL" 

\LPOP_SPEEDl.QUALITAnVE_VALUE\="NORMAL" 

\HPOP_BAL CAV_PRl.QUALITAnVE_VALUE\= w NORMAL" 

\LPFP_SPEEDl.QUALITAnVE VALUE\="LOW" 

\HI>OT_DS_TMPl.QUALrrATIVE_VALUE\="NORMAL" 

\HPFr_DSjrMPl.QUALrrATrVE_VALUE\= ,, NORMAL" 

\HPFP_SPEEDl.QUALrrATTVE_VALUE\=''NORMAL'' 

\^^FV_POSmON.QUALITATIVE_VALUE\= ,, NORMAL' , 

\MOV POSmON.QUALrTATIVE_VALUE\=^ORMAL" 

\CCVl>OSITION.QUALITAT^VE_VALUE\= >, NORMAL ,, 

********** 



302 Final Report, July 1993 



PBM numeric deviations 



A.5.2MCC Leak Example: Comparison of numerical data 

PBM numeric deviations 



Slice 29 


711*1 »1613 


Fil«2 *1614 


Diff 2-1 


\ 


1 


1.039888 


1.040192 


0.000303 


0.03 


2 


6.008003 


5.996758 


-0.011246 


-0.19 


3 


14.668247 


14.651808 


-0.016439 


-0.11 


4 


24.534515 


24.639374 


0.104858 


0.43 


5 


98.367554 


98.842560 


0.475006 


0.48 


6 


37.747559 


37.457550 


-0.29000* 


-0.77 


7 


165.911163 


166.015625 


0.104462 


0.06 


8 


0.229589 


0.231791 


0.002202 


0.96 


9 


1.690277 


1.879307 


0,189030 


11.18 


10 


3125.905273 


3126.816895 


0.911621 


0.03 


11 


0. 000000 


0.000000 ' y 


0.000000 




12 


1.000000 


1.000000 


0,000000 


0.00 


13 


1.000000 


1.000000 


0.000000 


0.00 


14 


1.000000 


1.000000 


0.000000 


0.00 


15 


26.500000 


26.500000 


0.000000 


0.00 


16 


15.299999 


15.299999 


0.000000 


0.00 


17 


0.013001 


0.013203 


0.000202 


1.56 


18 


0.013330 


0.013330 


. oooooo 


0.00 


19 


7500.000000 


7500.000000 


. oooooo 


0.00 


20 


1.000000 


1.000000 


o^oooooo 


0.00 


21 


0.000000 


0.000000 


0.000000 




22 


498.055176 


523.206543 


25.151367 


5.05 


23 


10.283998 


10.283998 


0.000000 


0.00 


24 


90.295990 


90.295990 


0.000000 


0.00 


25 


390254.125000 


390454.875000 


200.750000 


0.05 


26 


14.668247 


14.651808 


-0.016439 


-0.11 


27 


33.000000 


33.000000 


0.000000 


0.00 


28 


19.619995 


19.619995 


0.000000 


0.00 


29 


3.971998 


3.971998 


0,000000 


0.00 


30 


83.064148 


83.064148 


0.000000 


0.00 


31 


1.092199 


1.092199 


0.000000 


0.00 


32 


10.799999 


10.799999 


. oooooo 


0.00 


33 


1.386000 


1.386000 


0.000000 


0.00 


34 


2.896000 


2.896000 


0.000000 


0.00 


35 


1.011114 


1.010729 


-0.000385 


-0.04 


36 


12.000000 


12.000000 


0.000000 


0.00 


37 


12.000000 


12.000000 


o_._o_o_oo_oo 


0.00 


38 


7.400000 


7.400000 


0.000000 


0.00 


39 


10.189999 


10.189999 


. oooooo 


0.00 


40 


11.724998 


11.724998 


0.000000 


0.00 


41 


6.849998 


6.849998 


0.000000 


0.00 


42 


5.000000 


5.000000 


0.000000 


0.00 


43 


6.000000 


6.000000 


0.000000 


0.00 


44 


10.089998 


10.089998 


0.000000 


0.00 


45 


14.919998 


14.919998 


0.000000 


0.00 


46 


16.919983 


16.919983 


0.000000 


0.00 


47 


0.500000 


0.500000 


0.000000 


0.00 


48 


0.000000 


0.000000 


0.000000 




49 


0.060000 


0.060000 


0.000000 


0.00 


50 


96.038086 


96.511475 


0.473 389 


0.49 


51 


-0.109161 


0.289368 


0.398529 


-365.08 


52 


0.215000 


0.215000 


0.6~66000 


0.00 


53 


0.383000 


0.383000 


0.000000 


0.00 


54 


12.019999 


12.019999 


0.000000 


0.00 


55 


5.199999 


5.199999 


0.000000 


0.00 


56 


5.199999 


5.199999 


O.OTOOOO 


0.00 


57 


2.900000 


2.900000 


0.000000 


0.00 


58 


1.000000 


1.000000 


0. oooooo 


0.00 


59 


0.990341 


0.995184 


0.004842 


0.49 


60 


0.972574 


0.980250 


0.007676 


0.79 


61 


1.023678 


1.013735 


-0.009943 


-0.97 


62 


1.000000 


1.000000 


o .000600 


0.00 


63 


0.985196 


0.974770 


-0.610426 


-1.06 


64 


1.062466 


1.077913 


0.015448 


1.45 


65 


1.016024 


1.012693 


-0,603330 


-0.33 


66 


0.968910 


0.944607 


-0.024303 


-2.51 


67 


38000.000000 


38000.000000 


0.000000 


0.00 


68 


76.906433 


76.906464 


0.000031 


0.00 


69 


470000.000000 


470000.000000 


0.060600 


0.00 


70 


6388.164062 


6388.164062 


0.000000 


0.00 


71 


2002.959961 


2002.959961 


0. oooooo 


0.00 


72 


6388.164062 


6388.164062 


0.000000 


0.00 


73 


0.996604 


0.996605 


0.000000 


0.00 


74 


0.978000 


0.978000 


0.000000 


0.00 


75 


82.500000 


82.500000 


0.600060 


0.00 


76 


82.500000 


82.500000 


0.000000 


0.00 


77 


2.549999 


2.549999 


0.000060 


0.00 


78 


25.799988 


25.799988 


0.000000 


0.00 


79 


3125.905273 


3126.816895 


0.911621 


0.03 


80 


1.000000 


1.000000 


0.000000 


0.00 


81 


1.000000 


1.000000 


0.006600 


0.00 


82 


1.000000 


1.000000 


0.000000 


0.00 


83 


1.000000 


1.000000 


0.006000 


0.00 


84 


1.000000 


1.000000 


0.000000 


0.00 


85 


412.000000 


412.000000 


0.006000 


0.00 


86 


0.000000 


0.000000 


0. 000500 




87 


1.015055 


1.014589 


-0. 000465 


-0.05 



Final Report, July 1993 



303 



88 


1.03425B 


1.036100 


0.001842 


0.18 


89 


0.996876 


0.994775 


-0.002101 


-0.21 


90 


0.962178 


0.960224 


-0.001954 


-0.20 


91 


0.989823 


1.003180 


0.013357 


1.35 


92 


190.919983 


190.919983 


0.000000 


0.00 


93 


12500.000000 


12500.000000 


0.000000 


0*00 


94 


0.041826 


0.042803 


0.000977 


2.34 


95 


5087.000000 


5087.000000 


0.000000 


0.00 


96 


0.187000 


0.187000 


0.000000 


0.00 


97 


1631.000000 


1631.000000 


0.000000 


0.00 


98 


0.002151 


0.002151 


0.000000 


0.00 


99 


0.259972 


0.260048 


0.000076 


0.03 


100 


0.250000 


0.250000 


0.000000 


0.00 


101 


4010.000000 


4010.000000 


0.000000 


0.00 


102 


0.001337 


0.001337 


0.000000 


0.00 


103 


11.781599 


11.781599 


0.000000 


0.00 


104 


0.002201 


0.002201 


0.000000 


0.00 


105 . 


0.010520 


0.010520 


0.000000 


0.00 


106 ■ 


0.603247 


0.003247 


0.000000 


0.00 


107 ■ 


0.006517 


0.006517 


0.000000 


0.00 


10S : 


16000.000000 


16000.000000 


0.000000 


0.00 


109 ' 


0.054520 


0.054520 


0.000000 


0.00 


110 ; 


0.000690 


0.000690 


0.000000 


0.00 


111 . 


0.001946 


0.001946 


0.000000 


0.00 


112 


0.027540 


0.027540 


0.000000 


0.00 


113 « 


0.002632 


0.002632 


0.000000 


0.00 


114 : 


0.104100 


0.104100 


0.000000 


0.00 


115 


0.002100 


0.002100 


0.000000 


0.00 


Lie . 


98558.000000 


98558.000000 


0.000000 


0.00 


117 


1390.000000 


1390.000000 


0.000000 


0.00 


118 


0.050000 


0.050000 


0.000000 


0.00 


119 


11.320000 


11.320000 


0.000000 


0.00 


120 


4881.000000 


4881.000000 


0.000000 


0.00 


121 


0.008000 


0.008000 


0.000000 


0.00 


122 


0.999158 


0.998992 


-0.000167 


-0.02 


123 


0.000000 


0.000000 


0.000000 




124 


0.000000 


0.000000 


0.000000 




125 


0.000000 


0.000000 


0.000000 




126 - 


4.059999 


4.059999 


0. 000000 


0.00 


127 < 


30668.437500 


29211.523438 


-1456.914062 


-4.75 


128 


6000.000000 


6000.000000 


0.000000 


0.00 


129 • 


0.020760 


0.020760 


0.000000 


o.oo 


130 


0.091000 


0.091000 


0.000000 


0.00 


131 


0.010043 


0.010475 


0.000433 


4.31 


132 


0.057900 


0.057900 


0.000000 


o.oo 


133 


0.076200 


0.076200 


0.000000 


0.00 


134 


0.009725 


0.009510 


-0-000215 


-2.21 


135 


0.974578 


0.974743 


0.000165 


0.02 


13« 


0.059080 


0.059080 


0.000000 


0.00 


137 


0.481500 


0.461500 


0.000000 


0.00 


138 


0.005700 


0.005700 


0.000000 


0.00 


139 


0.997976 


1.076662 


0.078686 


7.88 


140 


1.199999 


1.199999 


0.000000 


0.00 


141 


0.047300 


0.047300 


0.000000 


0.00 


142 


1.879000 


1.879000 


0.000000 


0.00 


143 


0.577500 


0.577500 


0.000000 


0.00 


144 


166.000000 


166.000000 


0.000000 


0.00 


145 


0.038400 


0.038400 


0.000000 


0.00 


146 


1.860998 


1.860998 


0.000000 


0.00 


14? 


0.169000 


0.169000 


0.000000 


0.00 


148 


: 30.000000 


30.000000 


0.000000 


0.00 


149 


0.005687 


0.005687 


0.000000 


0.00 


150 


0.052300 


0.052300 


0.000000 


0.00 


151 


174600.000000 


174600.000000 


0. 000000 


0.00 


152 


0.000686 


0.000686 


0.000000 


0.00 


153 


0.352800 


0.352800 


0.000000 


0.00 


154 


766.000000 


766.000000 


0.000000 


0.00 


155 


0.134000 


0.134000 


0.000000 


0.00 


156 


0.112500 


0.112500 


0.000000 


0.00 


157 


: 0.192000 


0.192000 


0.000000 


o.oo 


158 


0.012600 


0.012600 


0.000000 


0.00 


159 


; 0.005000 


0.005000 


0.000000 


0.00 


160 


0.300000 


0.3 00000 


0.000000 


0.00 


161 


0.618500 


0.618500 


0.000000 


0.00 


162 


1.199999 


1.199999 


0.000000 


0.00 


163 


: 0.173400 


0.173400 


0.000000 


0.00 


164 


0.529000 


0.529000 


0.000000 


0.00 


165 


: 0.060347 


0.059739 


-0.000608 


-1.01 


166 


0.020810 


0.020810 


0.000000 


0,00 


167 


0.011800 


0.011800 


0.000000 


0.00 


168 


1.556257 


1.687683 


0.131426 


8.44 


169 


0.153000 


0. 153000 


0.000000 


0.00 


170 


138850.000000 


138850.000000 


0.000000 


0.00 


171 


74568.000000 


74568.000000 


0.000000 


0.00 


172 


339270.000000 


339270.000000 


0.000000 


0.00 


173 


89965.000000 


89965.000000 


0.000000 


0.00 


174 


45138.000000 


45138.000000 


0.000000 


0.00 


175 


535500.000000 


535500.000000 


o.oooooo 


0.00 


176 


30.000000 


30.000000 


0.000000 


0.00 


177 


130.000000 


130.000000 


0.000000 


0.00 


178 


: 245000.000000 


245000.000000 


0.000000 


0.00 


179 


194000.000000 


194000.000000 


0.000000 


0.00 


180 


: 0.007120 


0.007120 


0.006666 


0.00 


181 


0.535802 


0.535314 


-0.000489 


-0.09 


182 


: 1.500000 


1.500000 


0.000000 


0.00 


183 


0.168000 


0.168000 


0.000000 


0.00 


184 


0.000000 


0.000000 


0.000000 




185 


: 65000.000000 


65000.000000 


0.000000 


o.oo 



304 



Final Report, July 1993 



186 


: 1.013733 


1.013346 


-0.000387 


-0.04 


187 


: 0.012500 


0.012500 


0.000000 


0.00 


188 


: 0.026500 


0.026500 


0.000000 


0.00 


189 


: 1.000000 


1.000000 


0.000000 


0.00 


190 


: 1.000000 


1.000000 


0.000000 


0.00 


191 


: 1.000600 


1.000000 


0.000000 


0.00 


192 


: 1.000000 


1.000000 


0.009000 


0.00 


193 


: 0.891073 


0.900923 


0.009850 


1.11 


194 


: 0.807200 


0.807200 


0.000000 


0.00 


195 


1.033400 


1.033400 


0.000000 


o.oo 


196 


1.196472 


1.250399 


0.053926 


4.51 


197 


1.000000 


1.000000 


0.000000 


0.00 


198 


20900.000000 


20900.000000 


0.000000 


0.00 


199 


29.000000 


29.000000 


0.000000 


0.00 


200 


: 1.000000 


1.000000 


0.000000 


o.oo 


201 


82.240143 


82.240143 


0.000000 


0.00 


202 


1.504730 


1.505049 


0,000319 


0.02 


203 


0.951398 


0.951558 


0.000160 


0.02 


204 


1.961967 


1.961432 


-0.000536 


-0.03 


205 


2882.286621 


2804.322266 


-77.964355 


-2.70 


206 


4325.671675 


4332.679688 


7.007812 


0.16 


207 


287.210449 


287.554688 


0.344238 


0.12 


208 


4205.500000 


4208.031250 


2.531250 


0.06 


209 


1.866611 


1.866415 


-0.000196 


-0.01 


210 


52000.000000 


52000.000000 


0.000000 


0.00 


211 


1570.885254 


1528.712891 


-42.172363 


-2.68 


212 


1570.979980 


1528.801758 


-42.178223 


-2.68 


213 


: 7775.234375 


7775.156250 


-0.078125 


0.00 


214 


7689.757812 


7692.625000 


2.867188 


0.04 


215 


1884.011230 


1897.846680 


13.835449 


0.73 


216 


1684.228027 


1639.615234 


-44.612793 


-2.65 


217 


: 1684.228027 


1639.615234 


-44.612793 


-2.65 


218 


-811.693848 


-809.597656 


2.096191 


-0.26 


219 


-822.295410 


-811.830566 


10.464844 


-1.27 


220 


0.000000 


0.000000 


0.000000 




221 


0.000000 


0.000000 


0.000000 




222 


-823.186035 


-820.675781 


2.510254 


-0.30 


223 


-840.590820 


-828.914551 


11.676270 


-1.39 


224 


4001.000000 


4001.000000 


o.oooooo 


0.00 


225 


6001.000000 


6001.000000 


0.000000 


0.00 


226 


1684.228027 


1639.615234 


-44.612793 


-2.65 


227 


47.846252 


47.629211 


-0.21T041 


-0.45 


228 


21.433075 


21.282898 


-0.150177 


-0.70 


229 


0.824784 


0.824721 


-0.6150062 


-O.Ol 


230 


401.263184 


407.114258 


5.851074 


1.46 


231 


r 48.910950 


49.184845 


0.273895 


0.56 


232 


73.433044 


79.667969 


6.234924 


B.49 


233 


36.060303 


36.121460 


0.061157 


0.17 


234 


24.890167 


25.028778 


0.138611 


0.56 


235 


2980.142090 


2966.480469 


-13.661621 


-0.46 


236 


0.921638 


0.929231 


0.007593 


0.82 


237 


49.548981 


58.451569 


8.902588 


17.97 


238 


5.763311 


5.504456 


-0.258856 


-4.49 


239 


173.086273 


184.453278 


11.367004 


6.57 


240 


11.612564 


11.090994 


-0.521570 


-4.49 


241 


185.549164 


168.618042 


-16.931122 


-9.12 


242 


1095.488281 


1141.797363 


46.309082 


4.23 


243 


36.402039 


36.422180 


0.020142 


0.06 


244 


13.305758 


14.411146 


1.105389 


8.31 


245 


43.699738 


52.865021 


9.165283 


20.97 


246 


3736.286133 


3741.070312 


4.784180 


0.13 


247 


87.075806 


87.200745 


0.124939 


0.14 


248 


9.349928 


9.655035 


0.305107 


3.26 


249 


2175.591309 


2156.150391 


-19.440918 


-0.89 


250 


1.000000 


1.000000 


0.000000 


0.00 


251 


78.837372 


79.275269 


0.437897 


0.56 


252 


0.000000 


0.000000 


0.000000 




253 


22.119446 


23.950073 


i;i30«2Y 


8.28 


254 


94.923065 


95.041077 


0.II86T1 


0.12 


255 


14.023527 


13.872250 


-0.151278 


-1.08 


256 


46.359161 


48.412476 


2.053314 


4.43 


257 


8.487036 


8.881512 


0.394476 


4.65 


258 


164.326233 


167.903473 


3.5772T0 


2.18 


259 


91.522461 


90.676697 


-6:84576Y 


-0.92 


260 


952.055664 


930.703125 


-21.352539 


-2.24 


261 


598.803223 


598.086914 


-0.716309 


-0.12 


262 


391.079102 


391.999512 


0.920410 


0.24 


263 


69.289520 


69.397858 


0.108337 


0.16 


264 


1002.171387 


969.575195 


-32.596191 


-3.25 


265 


13.633186 


15.300093 


12.23 


266 


-6.989126 


16.691650 


uamn 


-338.82 


267 


937.333008 


796.952637 


-140.380X71 


-14.98 


268 


429.449707 
1037.450684 


422.855957 
1032.199219 


-€.593750 


-1.54 


269 


-5.251465 


-0.51 


270 


12.724550 


28.465485 


15.740934 


123.71 


271 


917.976074 


893.815918 


-24.160156 


-2.63 


272 


476.230957 


458.404785 


-17.826172 


-3.74 


273 « 


1193.688477 


1176.644531 


-17.043945 


-1.43 


274 . 


27.726288 


30.050476 


2.324188 


8.38 


275 . 


32.499420 


31.051636 


-1.447784 


-4.45 


276 : 


715.341797 


718.124023 


2.782227 


0.39 


277 : 


17.146332 


17.225677 


0.079346 


0.46 


278 : 


184.957184 


178.494598 


-€.462585 


-3.49 


279 : 


2818.295410 


2794.435059 


-23.860352 


-0.85 


280 : 


223.852417 


214.844696 


-9.007721 


-4.02 


281 : 


57.177246 


55.225739 


-1.951508 


-3.41 


282 : 


89.404022 


89.532318 


0.128296 


0.14 


283 : 


22.595673 


22.443237 


-0.152435 


-0.67 



Final Report, July 1993 



305 



284 : 


85000.000000 


85000.000000 


0.000000 


0.00 


285 : 


22.900360 


23.249603 


0.349243 


1.53 


286 : 


100.064026 


97.522186 


-2.541840 


-2,54 


287 : 


232.590607 


382.661621 


150.071014 


64.52 


288 : 


115.906189 


117.559448 


1.653259 


1.43 


289 : 


136.557861 


172.275635 


35.717773 


26.16 


290 : 


49.014557 


46.634979 


-6.379578 


-0.77 


291 : 


0.000000 


0.000000 


0. 000000 




292 : 


26.439606 


26.339050 


-0.100555 


-0.38 


293 : 


50.766113 


30.391418 


-20.374695 


-40.13 


294 : 


98.607452 


102.419006 


3.811554 


3.87 


295 : 


-29.285614 


-7.330830 


21.954784 


-74.97 


296 : 


436.204102 


387.557129 


-48.646973 


-11.15 


297 : 


1302.503418 


1425.779785 


123.2753T7 


9.46 


298 : 


-127.043030 


-120.938751 


6.104279 


-4.80 


299 : 


359.540039 


353.966309 


-5.573730 


-1.55 


300 : 


1429.694824 


1441.661816 


12.166992 


0.65 


301 : 


13.305758 


14.411146 


1.105389 


8.31 


302 ; 


54.147552 


69.767334 


15.619781 


28.65 


303 : 


126.626099 


126.671204 


0.045105 


0.04 


304 : 


330.485840 


332.415527 


1.929688 


0.58 


305 : 


15.696051 


-9.086666 


-24.782717 


-157.89 


306 : 


645.154785 


755.785645 


li<r.550T39 


17.15 


307 : 


16.421692 


18.499146 


2.077454 


12.65 


308 : 


12.676003 


12.870476 


0.194475 


1.53 


309 : 


276.560059 


267.552734 


-9.007324 


-3.26 


310 : 


3844.421875 


3849.672363 


5.250468 


0.14 


311 : 


6998.023438 


7013.085938 


15.062500 


0.22 


312 : 


5.715288 


5.701145 


-0.014143 


-0.25 


313 : 


-822.190430 


-820.067383 


2.123047 


-0.26 


314 : 


-840.696289 


-830.068848 


10.627441 


-1.26 


315 : 


0.000000 


0.000000 


0. 000000 




316 : 


0.488801 


0.488801 


0.000000 


0.00 


317 . 


5.342014 


5.157238 


-0.184776 


-3.46 


318 : 


54.164490 


53.446075 


-0.718414 


-1.33 


319 . 


17.381287 


16.578461 


-0.602826 


-4.62 


320 


3.092218 


-8.996962 


-12.OB9180 


-390.95 


321 


7.261265 


9.262882 


2.001617 


27.57 


322 ' 


150.469818 


149.748535 


-0.721283 


-0.48 


323 


112.326050 


111.126038 


-1.200012 


-1.07 


324 


-37.959106 


-37.727905 


0.231201 


-0.61 


325 


-44.444550 


-44.207581 


0.236969 


-0.53 


326 


-909.123535 


-908.835938 


0.287598 


-0.03 


327 


-48.000000 


-47.794037 


0.205963 


-0.43 


328 


-846.071289 


-834.468750 


11.602539 


-1.37 


329 


0.000000 


0.000000 


0.000000 




330 


1448.466797 


1405.748535 


-42". 7 18253 


-2.95 


331 


234954.000000 


233472.625000 


-1481.375000 


-0.63 


332 


0.000000 


0.000000 


o. oobboo 




333 


0,000000 


0.000000 


0.000000 




334 


356.607422 


357.732910 


1.125488 


0.32 


335 


381.881836 


371.177246 


-10.704590 


-2.80 


336 


24.035126 


24.049927 


0.014801 


0.06 


337 


11.282290 


11.287514 


0.005224 


0.05 


338 


2.179998 


2.179998 


0.000000 


0.00 


339 


0. 280000 


0. 280000 


0. booooo 


0.00 


340 


0.708144 


0.709683 


0.001539 


0.22 


341 


0.753538 


0.757118 


0.003580 


0.48 


342 


0.526800 


0.532320 


0.005521 


1.05 


343 


0.819946 


0.811429 


-0.008517 


-1.04 


344 


0.689397 


0.689415 


0.000019 


0.00 


345 


0.663454 


0. 656520 


-0.006933 


-1.05 


346 


0.254709 


0.254729 


0.000020 


0.01 


347 


136.755310 


-5.811361 


-142.566671 


-104.25 


348 


7364.046875 


6951.328125 


-412.718750 


-5.60 


349 


5565.273438 


5801.593750 


236.320312 


4.25 


350 


: 1.034258 


1.036100 


0.001842 


0.18 


351 


37.837921 


35.632721 


-2.205200 


-5.83 


352 


0.838997 


0.852369 


0.013372 


1.59 


353 


0.645903 


0.643710 


-0. 002193 


-0.34 


354 


: 0.743599 


0.725165 


-0.0IS4T4 


-2.46 


355 


: 363.520996 


363.593750 


0.072754 


0.02 


356 


: 363.636719 


363.709473 


0-.6T2734 


0.02 


357 


: 451.089355 


451.037109 


-0.052246 


-0.01 


358 


484108.500000 


464204.125000 


95.625000 


0.02 


359 


: 15866.257812 


15499.335938 


-366.921875 


-2.31 


360 


34826.398438 


34778.031250 


-48.367186 


-0.14 


361 


488801.500000 


488601.500000 


0. 000000 


0.00 


362 


: 390076.625000 


390172.125000 


95.500000 


0.02 


363 


: 0.000000 


0.000000 


0.000000 




364 


5222.015625 


5224.867188 


2.851562 


0.05 


365 


28395.640625 


28431.826125 


36.187500 


0.13 


366 


: 390254.125000 


390454.875000 


200.750000 


0.05 


367 


17.659180 


17.555420 


-0.103760 


-0.59 


368 


484108.500000 


484204.125000 


95.625000 


0.02 


369 


: 199.025238 


195.438293 


-3.586945 


-1.80 


370 


: 0.000000 


0.000000 


0. 000000 




371 


1582.302734 


1582.088379 


-0.214355 


-0.01 


372 


206.733154 


203.145294 


-3.587660 


-1.74 


373 


: 1584.300293 


1584.086426 


-0.213667 


-0.01 


374 


: 206.557373 


203.082489 


-3.474884 


-1.68 


375 


: 1888.576660 


1897.614746 


9.038086 


0.48 


376 


: -1.329216 


-1.329277 


-0.000061 


0.00 


377 


: -1.056177 


-1.056089 


0.000088 


-0.01 


378 


: 6780.390625 


8306.250000 


-474,140625 


-5.40 


379 


: 177046.750000 


176766.250000 


-280.500000 


-0.16 


380 


: -810.960449 


-808.601758 


2.158691 


-0.27 


381 


-785.354492 


-790.323730 


-4.969238 


0.63 



306 



Final Report, July 1993 



382 


: 846.370605 


846.716797 


0.346191 


0.04 


383 


: 871.657227 


872.535645 


0.878418 


0.10 


384 


: 681.120117 


680.488281 


-0.631836 


-0.09 


385 


: 8013.757812 


8025.851562 


12.093750 


0.15 


386 


: 6419.710938 


6446.195312 


26.484375 


0.41 


387 


: 3459.628418 


3272.562500 


-187.065918 


-5.41 


388 


: 65750.250000 


65477.937500 


-272.312500 


-0.41 


389 


3459.629395 


3272.563965 


-187.065430 


-5.41 


390 


: 65750.250000 


65477.937500 


-272.312500 


-0.41 


391 


: 1656.234375 


1655.021973 


-1.212402 


-0.07 


392 


24214.796875 


24512.671875 


297.875000 


1.23 


393 


1531.616699 


1545.909668 


14.292969 


0.93 


394 


1656.234375 


1655.021973 


-1.212402 


-0.07 


395 


: 25746.414062 


26058.578125 


312.1(4062 


1.21 


396 


: 141.935944 


138.685883 


-3.250061 


-2.29 


397 


-822.038574 


-811.555664 


10.482910 


-1.28 


398 


: 0.000000 


0.000000 


0.000000 




399 


: 822.769531 


823.019043 


0.249512 


0.03 


400 


-106.967682 


-107.662598 


-0.694916 


0.65 


401 


-86.957184 


-88.561920 


-1.604736 


1.85 


402 


1658.905762 


1612.898438 


M5. 0073 2 4 


-2.77 


403 


-535.488281 


-533.989258 


1.499023 


-0.28 


404 


-583.459473 


-578.058105 


5.401367 


-0.93 


405 


-91.035767 


-92.623718 


-1.587952 


1.74 


406 


214.939392 


211.431244 


-3.508148 


-1.63 


407 


1493.018066 


1455.863281 


-37.154785 


-2.49 


408 


1571.516113 


1529.305176 


-42.210938 


-2.69 


409 


-841.B49121 


-838.150391 


3.698730 


-0.44 


410 


-»5. 685577 


-96.989746 


-1.304169 


1.36 


411 


131.542969 


129.597870 


-1.945099 


-1.48 


412 


-28.012848 


-27.895447 


0.117401 


-0.42 


413 


-37.676910 


-37.445801 


0.231110 


-0.61 


414 


3358.910645 


3371.488770 


12.578125 


0.37 


415 


3383.432617 


3401.971680 


18.539062 


0.55 


416 


4258.562500 


4263.554688 


4.992188 


0.12 


417 


5673.867188 


5672.687500 


-1.179688 


-0.02 


418 


3442.862305 


3442.001953 


-0.866352 


-0.02 


419 


3410.314941 


3411.146973 


0.832031 


0.02 


420 


5114.882812 


5134.750000 


19.867188 


0.39 


421 


-54.758698 


-54.716400 


0.042297 


-0.08 


422 


-56.028168 


-55.984680 


0.043488 


-0.08 


423 


-53.197266 


-53.153717 


0.043549 


-0.08 


424 


-37.844604 


-37.612885 


0.231720 


-0.61 


425 


-936.622559 


-931.647949 


4.974609 


-0.53 


426 


3360.766113 


3352.695312 


-8.070801 


-0.24 


427 


0.229586 


0.234845 


0.005258 


2.29 


428 


0.185572 


0.185726 


0.000154 


0.08 


429 


0.213036 


0.213020 


-0. 000015 


-0.01 


430 


0.136324 


0.136216 


-0.000168 


-0.08 


431 


0.085357 


0.087292 


0.001935 


2.27 


432 


3136.913574 


3137.850586 


0.937012 


0.03 


433 


3383.800781 


3396.517578 


12.716797 


0.38 


434 


4200.765625 


4190.304688 


-10.4*0*38 


-0.25 


435 


3310.000000 


3322.303711 


12.303711 


0.37 


436 


3416.078125 


3416.651367 


0.573242 


0.02 


437 


3396.738770 


3416.382812 


19.644043 


0.58 


438 


5838.195312 


5840.593750 


2.398438 


0.04 


439 


4921.500000 


4865.007812 


-56.492188 


-1.15 


440 


3153.584473 


3154.542480 


0.958008 


0.03 


441 


3555.613281 


3434.521484 


-121.091797 


-3.41 


442 


3404.465820 


3405.560059 


1.094238 


0.03 


443 


5165.000000 


5173.625000 


8.625000 


0.17 


444 


1.358866 


1.345037 


-0.013828 


-1.02 


445 


1.472212 


1.472334 


0.000122 


0.01 


446 


3548.352051 


3425.258789 


-123.093262 


-3.47 


447 


5659.937500 


5646.625000 


-13.312500 


-0.24 


448 


5705.492188 


5705.609375 


0.117188 


0.00 


449 


41623.546875 


41633.492188 


9^45312 


0.02 


450 


0.001470 


0.001470 


07666660 


0.00 


451 : 


1.502628 


1.495081 


-0.007547 


-0.50 


452 


3084.205566 


3085.142090 


0.936523 


0.03 


453 : 


0.409330 


0.405777 


-0,003553 


-0.87 


454 


1.713087 


1.715132 


0.0020*5 


0.12 


455 : 


0.307038 


0.306419 


-0.660620 


-0.20 


456 , 


0.357950 


0.357579 


-6.0TO372 


-0.10 


457 : 


0.538200 


0.539045 


0.000846 


0.16 


458 : 


6186.007812 


6175.179688 


-10.828125 


-0.18 


459 : 


0.000000 


0.000000 


0.000000 




460 : 


256.779297 


242.852478 


-13.926819 


-5.42 


461 : 


4679.367188 


4631.289062 


-48.078125 


-1.03 


462 • 


5097.734375 


5117.523438 


19.789062 


0.39 


463 


4240.890625 


4246.195312 


5.304688 


0.13 


464 . 


396.470215 


396.530273 


0.060059 


0.02 


465 « 


4169.156250 


4174.015625 


4.859375 


0.12 


466 . 


5137.273438 


5143.570312 


6.296875 


0.12 


467 : 


23.598419 


23.701996 


0.103577 


0.44 


468 : 


229.530762 


215.567596 


-13.963165 


-6.08 


469 : 


4276.343750 


4281.351562 


570TT7812 


0.12 


470 : 


1383.277344 


1386.649902 


3.372559 


0.24 


471 : 


-55.152985 


-55.110321 


0.042664 


-0.08 


472 : 


-42.900238 


-42.841156 


0^059082 


-0.14 


473 : 


-29.595795 


-29.330048 


6.265747 


-0.90 


474 : 


5949.226562 


5938.125000 


-UU6i562 


-0.19 


475 : 


3978.882324 


3982.648926 


3.766602 


0.09 


476 : 


7372.367188 


7385.437500 


13.070312 


0.18 


477 : 


5594.445312 


5596.476562 


2.031250 


0.04 


478 : 


6202.445312 


6205.820312 


3.375000 


0.05 


479 : 


6903.703125 


6986.960938 


837*57812 


1.21 



Final Report, July 1993 



C'L 



307 



480 : 


6904.117188 


6930.578125 


26.460938 


0.38 


481 : 


5591.109375 


5593.156250 


2.046875 


0.04 


482 : 


6308.570312 


6311.398438 


2.828125 


0.04 


483 : 


4182.734375 


4188.382812 


5.648439 


0.14 


484 : 


4254.914062 


4260.070312 


5.156250 


0.12 


485 : 


292.839355 


278.973633 


-13.865723 


-4.73 


486 : 


6280.929688 


6270.226562 


-10.703125 


-0.17 


487 : 


432.872070 


432.952637 


0.080566 


0.02 


488 : 


4345.632812 


4350.750000 


5.117188 


0.12 


489 : 


7394.492188 


7409.617188 


15.125000 


0.20 


490 : 


265.698242 


251.788696 


-13.909546 


-5.24 


491 : 


6044.148438 


6033.164062 


-10.984375 


-0.18 


492 : 


3546.263184 


3424.866699 


-121.396484 


-3.42 


493 : 


4228.500000 


4233.726562 


5,226562 


0.12 


494 : 


6438.210938 


6592.070312 


153.859375 


2.39 


495 : 


6417.531250 


6455.671875 


38.140625 


0.59 


496 : 


0.000000 


0.000000 


0.000000 




497 : 


4071.387207 


4076.073730 


4.686523 


0.12 


498 : 


5805.945312 


5792.664062 


-13.281250 


-0.23 


499 : 


0.000000 


0.000000 


0.000000" 




500 : 


0. 000000 


0.000000 


0.000000 




501 : 


3448.577637 


3447.702637 


-0.875000 


-0.03 


502 : 


3462.638184 


3475.792480 


13.154297 


0.38 


503 : 


356.441406 


356.473633 


0.032227 


0.01 


504 : 


3418.857910 


3440.333008 


21.475098 


0.63 


505 : 


5725.976562 


5712.328125 


-13.648438 


-0.24 


506 : 


4099.335938 


4104.078125 


4.742188 


0.12 


507 : 


7361.257812 


7374.835938 


13.578125 


0.18 


508 : 


3852.255371 


3855.974609 


3.719238 


0.10 


509 : 


4001.099609 


4001.099609 


0.000000 


0.00 


510 : 


12008.195312 


707.099609 


-11301.095703 


-94.11 


511 : 


90503.000000 


90503.000000 


0.000000 


0.00 


512 : 


2417.099609 


2417.099609 


0.000000 


0.00 


513 J 


4864.320312 


4809.781250 


-54.539062 


-1.12 


514 : 


1.554476 


1.573479 


0.019003 


1.22 


515 : 


6139.64B438 


6126.773438 


-12.875000 


-0.21 


516 : 


4.683475 


4.708469 


0.024994 


0.53 


517 : 


4.695236 


4.717484 


0.022247 


0.47 


518 : 


7363.101562 


7377.492188 


14.390625 


0.20 


519 : 


6016.984375 


6006.804688 


-10.179688 


-0.17 


520 . 


6435.039062 


6588.484375 


153.445312 


2.38 


521 


6023.742188 


6009.210938 


-14.531250 


-0.24 


522 


6445.125000 


6483.671875 


38.546875 


0.60 


523 


5676.960938 


5663.695312 


-13.265625 


-0.23 


524 


3526.417969 


3514.186035 


-12.231934 


-0.35 


525 


5603.835938 


5621.625000 


17.789062 


0.32 


526 


7386.007812 


7400.734375 


14.726562 


0.20 


527 . 


1.139999 


1.139999 


o.oooooo 


0.00 


528 . 


5587.453125 


5613. 171875 


25.718750 


0.46 


529 


6467.500000 


6599.406250 


131.906250 


2.04 


530 


2.506416 


2.512877 


0.006460 


0.26 


531 


2.522751 


2.525522 


0.002771 


0.11 


532 


6544.578125 


6576.617188 


32.039062 


0.49 


533 


68.597870 


68.594818 


-0.003052 


0.00 


534 


68.654388 


68.698822 


0.044434 


0.06 


535 


4136.976562 


4143.04687S 


6.070312 


0.15 


536 


3735.716797 


3735.937500 


0.220703 


0.01 


537 


€117.054688 


6104.328125 


-12.726562 


-0.21 


538 


0.000000 


0.000000 


0.000000 




539 


5630.273438 


5647.960938 


17.687500 


0.31 


540 


5686.062500 


5715.593750 


29.531250 


0.52 


541 


5658.171875 


5681.773438 


23.601562 


0.42 


542 


15753.796875 


15741.945312 


-11.851562 


-0.08 


543 


16137.304688 


16128.265625 


-9.039062 


-0.06 


544 


5854.031250 


5856.804688 


2.773438 


0.05 


545 


7075.421875 


7078.796875 


3.375000 


0.05 


546 


69.766113 


69.778046 


0.011932 


0.02 


547 


69.788757 


69.951843 


0.163086 


0.23 


548 


70.401031 


70.472748 


0.071716 


0.10 


549 


6.008545 


6.014065 


0.005520 


0.09 


550 


0.000000 


0.000000 


0.000000 




551 


0.763860 


0.782395 


0.018535 


2.43 


552 


0.896606 


0.904724 


0.008117 


0.91 


553 


0.798541 


0.816353 


0.017812 


2.23 


554 


0.000000 


0.000000 


0.000000 




555 


654.646484 


670.342773 


15.696289 


2.40 


556 


6.009394 


5.996889 


-0.012505 


-0.21 


557 


: 6.008003 


5.99*758 


-0.011246 


-0.19 


55S 


0.943377 


0.951450 


0.008074 


0.86 


559 


: 0.673226 


0.699126 


0.025900 


3.85 


560 


0.658929 


0.684588 


0.025659 


3.89 


561 


: 4.372171 


4.384647 


0.012476 


0.29 


562 


70.689941 


70.672211 


-0.017731 


-0.03 


563 


: 4.372171 


4.384647 


0.012476 


0.29 


564 


: 4.281281 


4.292648 


0.011368 


0.27 


565 


69.943634 


69.925110 


-0.018524 


-0.03 


566 


: 4.305275 


4.316406 


0.011131 


0.26 


567 


4.873102 


4.889681 


0.016579 


0.34 


568 


: 4.873102 


4.889681 


0.016579 


0.34 


569 


69.949493 


69.871460 


-0.078033 


-0.11 


570 


: 70.389709 


70.385864 


-0.003845 


-0.01 


571 


0.642401 


0,642843 


0.000443 


0.07 


572 


; 0.524147 


0.537056 


0.012909 


2.46 


573 


: 0.636804 


0.675765 


0.038961 


6.12 


574 


: 1.824329 


1,824579 


0.000250 


0.01 


575 


: 2.700668 


2.701349 


0.000681 


0.03 


576 


: 71.013275 


70.993683 


-0.019592 


-0.03 


577 


: 4.836267 


4.852570 


0.016302 


0.34 



308 



Final Report, July 1993 



w 



i ^Sl. 



578 


97.309540 


91.888000 


-5.421539 


-5.57 


57 9 


12.124474 


12.073769 


-0.050705 


-0.42 


580 


0.000000 


0.000000 


0.000000 




581 


681.960449 


664.615234 


-17.345215 


-2.54 


582 


4.801439 


4.817924 


0.016485 


0.34 


583 


97.254059 


96.060883 


-1.193176 


-1.23 


584 


0.000000 


0.000000 


0.000000 




585 


277.473145 


277.477539 


0.004395 


0.00 


58 € 


98.215729 


97.020081 


-1.195648 


-1.22 


587 


492.625000 


480.726562 


-11.898438 


-2.42 


588 


492.624512 


480.726562 


-11.897949 


-2.42 


589 


0.000000 


0.000000 


0.000000 




590 


1708.430176 


1721.697754 


13.267578 


0.78 


591 


1577.485840 


1602.368652 


24.882812 


1.58 


592 


1334.388672 


1379.204102 


44.815430 


3.36 


593 


0.000000 


0.000000 


0.000000 




594 


454.350098 


454.282715 


-0.067383 


-0.01 


595 


846.847656 


846.457520 


-0.390137 


-0.05 


596 


171.714905 


171.821136 


0.106232 


0.06 


597 


1732.790527 


1744.819336 


12.028809 


0.69 


598 


184.205322 


184.700958 


0.495636 


0.27 


599 


1313.631348 


1354.135742 


40.504395 


3.08 


600 


0.000000 


0.000000 


0.000000 




601 


363.355469 


363.330566 


-0.024902 


-0.01 


602 


363.471191 


363.446289 


-0.024902 


-0.01 


603 


492.625000 


480.726562 


-11.898438 


-2.42 


604 


1145.219238 


1108.941895 


-36.277344 


-3.17 


605 


9915.687500 


9888.343750 


-27.343750 


-0.28 


606 


1665.778809 


1663.651367 


-2.127441 


-0.13 


607 


4762.101562 


4813.703125 


51.601562 


1.08 


608 


1918.015625 


1931.161621 


13.145996 


0.69 


609 


1465.938477 


1510.327148 


44.388672 


3.03 


610 


734.142578 


725.390137 


-8.752441 


-1.19 


611 


547.169922 


550.911621 


3.741699 


0.68 


612 


0.000000 


0.000000 


0.000000 




613 


277.473145 


277.477539 


0.004395 


0.00 


614 


0.000000 


0.000000 


0.000000 




615 


43.089569 


42.614777 


-0.474792 


-1.10 


616 


479.625000 


467.726562 


-11.898438 


-2.48 


617 


0.475666 


0.477139 


0.001473 


0.31 


618 


0.578000 


0.578000 


0.000000 


0.00 


619 


0.156009 


0.155901 


-0.000108 


-0.07 


620 


: 0.000000 


0.000000 


0.000000 




621 


41.626526 


41.138184 


-0.488342 


-1.17 


622 


166.781250 


166.885895 


0.104645 


0.06 


623 


181.745239 


181.876404 


0.131165 


0.07 


624 


203.345856 


203.989136 


0.643280 


0.32 


625 


0.000000 


0.000000 


0.000000 




626 


97.254059 


96.060883 


-1.193176 


-1.23 


627 


462.243652 


451.148438 


-11.095215 


-2.40 


628 


1767.545898 


1781.413086 


13.867188 


0.78 


629 


1729.335938 


1742.803711 


13.467773 


0.78 


630 


39.232086 


38.839783 


-0.392303 


-1.00 


631 


67.356079 


66.629120 


-0,726959 


-1.08 


632 


1259.465332 


1300.248047 


40.782715 


3.24 


633 


167.752167 


167.855713 


0.103546 


0.06 


634 


195.293732 


195.871735 


0.578003 


0.30 


635 


207.586731 


207.833557 


0.246826 


0.12 


636 


192.206604 


192.328003 


0.121399 


0.06 


637 


1353.612305 


1399.201172 


45.588867 


3.37 


638 


1342.388672 


1387.204102 


44.815430 


3.34 


639 


97.742859 


96.549683 


-1.193176 


-1.22 


640 


0.177740 


0.178457 


0.000717 


0.40 


641 


0.357969 


0.356894 


-0.001075 


-0.30 


642 


0.475998 


0.475688 


-0.000310 


-0.07 


643 


0.297264 


0.297463 


0.000199 


0.07 


644 


830.754395 


811.542480 


-19.211914 


-2.31 


645 


1823.504395 


1820.972168 


-2.532*27 


-0.14 


646 


0.132820 


0.104495 


-0.028325 


-21.33 


647 


0.670025 


0.698373 


0.028348 


4.23 


648 


0.041146 


0.041231 


0.000085 


0.21 


649 


211.297852 


211.459167 


0.161316 


0.0B 


650 


211.234375 


210.968689 


-0.265686 


-0.13 


651 


512.298340 


500.451172 


-11.847168 


-2.31 


652 


1548.458984 


1546.308594 


-2.150391 


-0.14 


653 


267.157715 


267.303711 


0.145996 


0.05 


654 


848.709961 


849.791504 


1.081543 


0.13 


655 


619.496094 


620.285645 


0.789551 


0.13 


656 


136.711914 


136.786591 


0.074677 


0.05 


657 


1250. 143555 


1251.736328 


1.592773 


0.13 


658 


450.945801 


450.893555 


-6.052246 


-0.01 


659 


18.288666 


18.834991 


0.546326 


2.99 


660 


3.928347 


3.921061 


-0.007286 


-0.19 


661 


153.461548 


153-783661 


0.322113 


0.21 


662 


0.833337 


0.833916 


0,000579 


0.07 


663 


27.977386 


27.666656 


-0.310730 


-1.11 


664 


0.000000 


0.000000 


0.000000 




665 


149.985504 


150.327637 


0.342133 


0.23 


666 


1.123104 


1.111996 


-0.011108 


-0.99 


667 


1076.307129 


1076.619141 


0*312012 


0.03 


668 


1.919867 


2.111099 


0.191233 


9.96 


669 


0.134956 


0.134993 


0.000037 


0.03 


670 


72.821747 


72.856537 


0.034790 


0.05 


671 


180.628387 


180.868530 


0.240143 


0.13 


672 


65.853943 


67.376312 


1.522369 


2.31 


673 


0.000000 


0.000000 


0.000000 




674 


0.827072 


0.810597 


-0.016475 


-1.99 


675 


0.433507 


0.435336 


0.001829 


0.42 



m 



Final Report, July 1993 



309 



676 


: 15.776806 


15.588564 


-0.188242 


-1.19 


677 


: 78,045044 


77.959869 


-0.085175 


-0.11 


678 


: 0.000000 


0.000000 


0.000000 




679 


: 73.625885 


74.174957 


0.549072 


0.75 


680 


: 153.461548 


153.783661 


0.322113 


0.21 


681 


: 153.913849 


154.248718 


0.334869 


0.22 


662 


: 1663.406250 


1617.179688 


-46.226562 


-2.78 


683 


: 153.108459 


153.430237 


0.321777 


0.21 


684 


: 155.599304 


156.055878 


0.456573 


0.29 


685 


: 0.123504 


0.121655 


-0.001849 


-1.50 


686 


: 27.977386 


27.666656 


-0.310730 


-1.11 


687 


: 151.670959 


152*134827 


0.463867 


0.31 


688 


230.525421 


232.698730 


2.173309 


0.94 


689 


27.295502 


26.969818 


-0.325684 


-1.19 


690 


: 3.826061 


3.841618 


0.015556 


0.41 


691 


3.346209 


3.397953 


0.051744 


1.55 


692 


818.146484 


815.833984 


-2.312500 


-0.28 


693 


3.917152 


3.933422 


0.016270 


0.42 


694 


3.359634 


3.410852 


0.051218 


1.52 


695 


3.614697 


3.650373 


0.035677 


0.99 


696 


1.063646 


1.139660 


0.076014 


7.15 


697 


0.626629 


0.739648 


0.113019 


18.04 


698 


0.190139 


0.190067 


-0.000072 


-0.04 


699 


0.345198 


0.346011 


0.000813 


0.24 


700 


1.950249 


2.139357 


0.189108 


9.70 


701 


921.997559 


922.203125 


0.205566 


0.02 


702 


0.535337 


0.536077 


0.000741 


0.14 


703 


817.611328 


815.297852 


-2.313477 


-0.28 


704 


0.134956 


0.134993 


0.000037 


0.03 


705 


11.518705 


11.381269 


-0.137436 


-1.19 


706 


0.000000 


0.000000 


0.000000 




707 


42.190735 


43.033905 


0.843170 


2.00 


708 


42.190735 


43.033905 


0.843170 


2.00 


709 


28.403870 


30.086121 


1.682251 


5.92 


710 


4.083572 


4.110640 


0.027067 


0.66 


711 


922.257324 


922.463379 


0.206055 


0.02 


712 


1102.733337 


1102.968262 


0.234375 


0.02 


713 


1102.598633 


1102.833496 


6.234663 


0.02 


714 


1110.665527 


1110.905273 


0.239746 


0.02 


715 


110.092377 


112.426971 


2.334595 


2.12 


716 


0.000000 


0.000000 


0.000000 




717 


920.088867 


920.104004 


0.015137 


0.00 


718 


71.300720 


73.824615 


2.523895 


3.54 


719 


0.083146 


0.085052 


0.001906 


2.29 


720 


180.476318 


180.505005 


0.028687 


0.02 


721 


0.816395 


0.805320 


-0.011075 


-1.36 


722 


0.519884 


0.544207 


0.024323 


4.68 


723 


112.042633 


114.566345 


2.523712 


2.25 


724 


818.146484 


815.833984 


-2.312500 


-0.28 


725 


0.000000 


0.000000 


0. 000000 




726 


4386.523438 


4352.312500 


-34.210938 


-0.78 


727 


6.130547 


6.134830 


0.004284 


0.07 


728 


1.936363 


1.936960 


0.000597 


0.03 


729 


69.285187 


71.821167 


2.535980 


3.66 


730 


0.452308 


0.465040 


0.012732 


2.81 


731 


102.025452 


104.355194 


2.329742 


2.28 


732 


4.256339 


4.204613 


-0.053726 


-1.26 


733 


3.625399 


2.818214 


-0.807184 


-22.26 


734 


0.373050 


0.377779 


0.004729 


1.27 


735 


0.617861 


0.614347 


-0.003513 


-0.57 


736 


0.912729 


0.911265 


-0.001464 


-0.16 


737 


1073.197754 


1073.534180 


0.336426 


0.03 


738 


119.531281 


120.211151 


0.679871 


0.57 


739 


62.674286 


62.853668 


0.179382 


0.29 


740 


56.856995 


57.357483 


0.500488 


0.88 


741 


1075.458984 


1075.986816 


0.527832 


0.05 


742 


1.368673 


1.369993 


0.001320 


0.10 


743 


0.206650 


0.206707 


0.000057 


0.03 


744 . 


930.189453 


930.399902 


0.210449 


0.02 


745 


27.884003 


29.541901 


1.657898 


5.95 


746 . 


73.192383 


73.739624 


0.547241 


0.75 


747 : 


0.000000 


0.000000 


0.000000 




748 : 


27.762146 


29.480896 


1.718750 


6.19 


749 : 


72.877106 


73.432007 


0.554901 


0.76 


750 : 


-0.121855 


-0.060998 


0.060657 


-49.94 


751 : 


0.641739 


0.605205 


-0.036534 


-5.69 


752 : 


-0.315278 


-0.307602 


0.007676 


-2.43 


753 : 


0.748785 


0.742938 


-0.005847 


-0.78 


754 : 


0.445283 


0.443784 


-0.001500 


-0.34 


755 : 


0.501109 


0.484511 


-0.016598 


-3.31 


756 : 


1.224766 


1.227062 


0.002296 


0.19 


757 : 


1.057533 


1.057219 


-0.000315 


-0.03 


758 : 


1.082760 


1.079916 


-0.002844 


-0.26 


759 : 


0.718942 


0.719197 


0.000255 


0.04 


760 : 


0.731064 


0.731316 


0.000253 


0.03 


761 J 


0.759342 


0.760455 


0.001113 


0.15 


762 : 


7.497343 


7.478819 


-0.018524 


-0.25 


763 : 


24.559753 


24.694458 


0.134705 


0.55 


764 : 


8.061710 


8.006901 


-0.054810 


-0.68 


765 : 


8.066910 


8.071793 


0.004883 


0.06 


766 : 


42.287079 


43.139862 


0.852783 


2.02 


767 ; 


77.217987 


77.149261 


-0.068726 


-0.09 


768 : 


25.763000 


27.443329 


1.680328 


6.52 


769 : 


70.232605 


70.778931 


0.546326 


0.78 


770 : 


41.374319 


42.238*07 


0.854279 


2.06 


771 : 


77.217987 


77.149261 


-0.068726 


-0.09 


772 : 


93.128479 


92.970154 


-0.158325 


-0.17 


773 : 


765.688965 


765.738770 


0.049805 


0.01 



310 



Final Report, July 1993 



774 


: 11603.078125 


11625.078125 


17.000000 


0.15 


775 


0.998785 


0.998579 


-0.000206 


-0.02 


776 


193.631470 


191.544281 


-2.087189 


-1.08 


777 


4.065384 


4.018787 


-0.046597 


-1.15 


77B 


407.959473 


406.846191 


-1.113281 


-0.27 


779 


0.000000 


0.000000 


6.000000 




780 


349.586914 


347.508301 


-2.078613 


-0.59 


781 


4.132656 


4.298428 


0,165771 


4.01 


782 


9.308701 


9.223932 


-0.084768 


-0.91 


783 


441.669922 


432.037109 


-9.632812 


-2.18 


784 


: -1.056177 


-1.056089 


0.000088 


-0.01 


785 


-1.056177 


-1.056089 


0.000088 


-0.01 


786 


80.359070 


80.729492 


0.370422 


0.46 


787 


6.554482 


7.454908 


0.900427 


13.74 


78B 


371.910156 


371.835938 


-0.074219 


-0.02 


789 


218.941498 


207.219727 


-11.721771 


-5.35 


790 


44.541626 


44.508118 


-0.033508 


-0.08 


791 


243.789093 


243.652557 


-0.136536 


-0.06 


792 


670,997070 


670.247559 


-0.749512 


-0.11 


793 


10.000000 


10.000000 


o.ooooloo 


0.00 


794 


0.000000 


0.000000 


0.000000 




795 


0.997315 


0.995551 


-0.001763 


-0.1B 


796 


1.003361 


1.004101 


0.000740 


0.07 


797 


0.995481 


0.996269 


0.000787 


0.08 


798 


0.674211 


0.68012B 


0.005917' 


0.B8 


799 


0.790184 


0.790598 


0.000414 


0.05 


800 


-583.140625 


-577.716309 


5.424316 


-0.93 


801 


195.293732 


195.871735 


0,578003 


0.30 


802 


0.111000 


0.111000 


. o 00000 


0.00 


803 


0.012140 


0.012140 


0,000000 


0.00 


804 


0.012140 


0.012140 


6 . 660000 


0.00 


805 


2.000000 


2.000000 


. 000000 


0.00 


806 


2.000000 


2.000000 


0.000000 


0.00 


807 


56.856995 


57.357483 


0.500488 


0.88 


808 


119.531281 


120.211151 


0.679871 


0.57 


809 


0.036645 


0.036531 


-0.000114 


-0.31 


810 


1.856365 


1.857166 


0.000801 


0.04 


811 


154.374847 


154.833710 


0.458862 


0.30 


812 


1713.430176 


1726.697754 


13.267578 


0.77 


813 


70.594635 


73.120056 


2.525421 


3.58 


814 


1746.141602 


1733.294434 


-12.847168 


-0.74 


815 


875.000000 


875.000000 


0.000000 


0.00 


816 


1.309431 


1.298895 


-0,01PS36 


-0.80 


817 


0.657000 


0.657000 


6.000000 


6.06 


818 


0.910189 


0.918422 


6 . 008234 


0.90 


819 


2818.295410 


2794.435059 


-23.860352 


-0.85 


820 


2897.132812 


2873.709961 


-23.422852 


-0.81 


821 


1630.719727 


1630.719727 


. 000000 


O.OO 


822 


7952.054688 


7952.054688 


0.000000 


0.00 


823 


2.703897 


2.698883 


-0.005014 


-0.19 


824 


1.224447 


1.222176 


-0.002272 


-0.19 


825 


0.000000 


0.000000 


0.000000 




826 


1.000000 


1.000000 


0.000000 


0.00 


827 


-819.878906 


-817.603027 


2.275879 


-0.28 


828 


3.853502 


3.869274 


6.0T5772 


0.41 


829 


523.574707 


527.900879 


4.326172 


0.83 


830 


0.000000 


0.000000 


o; 000000 




831 


275.510742 


275.036133 


-0.474609 


-0.17 


832 


6.157999 


6.157999 


0,000000 


0.00 


833 


69.395416 


69.318451 


-0.07*965 


-0.11 


834 


3576.249512 


3584.871094 


B.62i5S2 


0.24 


835 


0.000000 


0.000000 


0.600000 




836 


0.000000 


0.000000 


6.000000 




837 


0.000000 


0.000000 


0.000000 




838 


0.000000 


0.000000 


0.000000 




839 


0.000000 


0.000000 


0.000000 




840 


0.000000 


0.000000 


0.000000 




841 


1200.000000 


1200.000000 


OoTOo 


6.06 


842 


0.000000 


0.000000 " 


67060000 




843 


365.000000 


365.000000 


o.oobooo 


0.00 


844 


385.000000 


385.000000 


0.000000 


0.00 


845 


330.000000 


330.000000 


0.000000 


0.00 


84C 


0.350000 


0.350000 


0.000000 


0.00 


847 


630.000000 


630.000000 


hmm 


6.60 


848 


0.000000 


0.000000 


0.000000 




849 


922.736328 


923.124512 


0.388184 


0.04 


850 


153.570709 


153.494293 


-0.076416 


-0.05 


851 : 


0.000000 


0.000000 


0,000000 




852 : 


0.000000 


0.000000 


o; 000000 




853 : 


0.000000 


0.000000 


6.000000 




854 - 


0.000000 


0.000000 


0,000000 




855 , 


0.000000 


0.000000 


0,000000 




856 


0.000000 


0.000000 


0^000000 




857 


0.000000 


0.000000 


0.000000 




858 . 


0.000000 


0.000000 


0.000000 




859 


0.000000 


0.000000 


0.000000 




860 : 


0.000000 


0.000000 


6^100000 




861 • 


0.000000 


0. 000000 


0.060666 




862 : 


0.000000 


0.000000 


0.000000 




863 : 


0.000000 


0.000000 


0.000000 




864 : 


0.000000 


0.000000 


0.000000 




865 : 


0.000000 


0.000000 


0.000000 




866 : 


0.000000 


0.000000 


o.oooodo 




867 : 


0.000000 


0.000000 


o". 000000 




868 : 


2034.550293 


1895.018066 


-139^532227 


-6.86 


869 : 


1.146242 


1.142563 


-0.003679 


-0.32 


870 : 


27.977386 


27.666656 


-0.3 107 30 


-1. 11 


871 : 


1.371799 


1.369883 


-6.6619ir 


-6.14 



Final Report, July 1993 



311 



872 


: 5551.804688 


5527.062500 


-24.T42188 


-0.45 


873 


: 1.510134 


1.502941 


-0.007193 


-0.48 


874 


: 0.857139 


0.854705 


-0.002434 


-0.28 


875 


26.159821 


26.461853 


0.302032 


1.15 


876 


: 1.350374 


1.349695 


-0.000679 


-0.05 


877 


: 6525.734375 


6509.585938 


-16.148438 


-0.25 


878 


1.484552 


1.484858 


0.000305 


0.02 


87* 


2.381550 


2.382854 


0.001305 


0.05 


880 


; 28.481201 


28.543182 


0.061981 


0.22 


881 


0.794742 


0.827302 


0.032560 


4.10 


882 


: 1.000000 


1.000000 


0.000000 


0.00 


883 


: 1.000000 


1.000000 


0.000000 


0.00 


884 


0.992441 


0.992410 


-0.000031 


0.00 


885 


1.060556 


1.064121 


0.003565 


0.34 


866 


1.030588 


0.817559 


-0.213029 


-20.67 


887 


: 1.090010 


1.049099 


-0.040911 


-3.75 


888 


0.994714 


0.994709 


-0.000005 


0.00 


889 


0.991223 


0.991211 


-0.000012 


0.00 


890 


3006.000000 


3006.000000 


0.000000 


0.00 


891 


0.000000 


0.000000 


0.000000 




892 


283.290527 


285.570312 


2.279785 


0.80 


893 


18.008453 


18.113037 


0.104584 


0.58 


894 


163.696625 


164.492493 


0.795868 


0.49 


895 


8730.429688 


8705.546875 


-24.882812 


-0.29 


896 


1.000000 


1.000000 


0.000000 


0.00 


897 


17.980042 


17.184479 


-0.795563 


-4.42 


898 


215.875580 


244.833008 


28.957428 


13.41 


899 


35360.187500 


31418.812500 


-3941.375000 


-11.15 


900 


1.000000 


1.000000 


0.000000 


0.00 


901 


3125.905273 


3126.816895 


0.911621 


0.03 


902 


5949.226562 


5938.125000 


-11.101562 


-0.19 


903 


4071.387207 


4076.073730 


4.686523 


0.12 


904 


153.461548 


153.783661 


0.322113 


0.21 


905 


921.997559 


922.203125 


0.205566 


0.02 


906 


237.616943 


223.666229 


-13.950714 


-5.87 


907 


43.089569 


42.614777 


-0.474792 


-1.10 


908 


15866.257812 


15499.335938 


-366.921875 


-2.31 


909 


34826.398438 


34778.031250 


-48.367188 


-0.14 


910 


1729.335938 


1742.803711 


13.467773 


0.78 


911 


5114.882812 


5134.750000 


19.867188 


0.39 


912 


356.441406 


356.473633 


0.032227 


0.01 


913 


5222.015625 


5224.867188 


2.851562 


0.05 


914 


195.293732 


195.871735 


0.578003 


0.30 


915 


7361.257812 


7374.835938 


13.578125 


0.18 


916 


28395.640625 


28431.828125 


36.187500 


0.13 


917 


1342.388672 


1387.204102 


44.815430 


3.34 


918 


5165.000000 


5173.625000 


8.625000 


0.17 


919 


3310.000000 


3322.303711 


12.303711 


0.37 


920 


479.625000 


467.726562 


-11.898438 


-2.48 


921 


0.000000 


0.000000 


0.000000 




922 


4626.335938 


4589.796875 


-36.539062 


-0.79 


923 


922.736328 


923.124512 


0.388184 


0.04 


924 


153.570709 


153.494293 


-0.076416 


-0.05 


925 


390254.125000 


390454.875000 


200.750000 


0.05 


926 


6.008545 


6.014065 


0. 005520 


0.09 


927 


1.040003 


1. 040003 


0.000000 


0.00 


928 


0.997315 


0.995551 


-0.001763 


-0.18 


929 


1.003361 


1.004101 


0.000740 


0.07 


930 


0.995481 


0.996269 


0.000787 


0.08 


931 


0.674211 


0.680128 


0.005917 


0.88 


932 


0.790184 


0.790598 


0.000414 


0.05 


933 


9010612.000000 


9010614.000000 


2.000000 


0.00 


934 


120889.000000 


121589.000000 


700.000000 


0.58 


935 


520.000000 


520.000000 


0.000000 


0.00 


936 


10.000000 


10.000000 


0.000000 


0.00 


937 


395.000000 


395.000000 


0.000000 


0.00 


938 


0. 000000 


0.000000 


0.000000 




939 


14.668247 


14.651808 


-0.016439 


-0.11 


940 


498.055176 


523.206543 


25.151367 


5.05 


941 


0.000000 


0.000000 


0.000000 




942 


10.283998 


10.283998 


0.000000 


0.00 


943 


90.295990 


90.295990 


0.000000 


0.00 


944 


47.500000 


47.500000 


0.000000 


0.00 


945 


47.500000 


47.500000 


0.000000 


0.00 


946 


0.000000 


0.000000 


0.000000 




947 


0.000000 


0.000000 


0.000000 




948 


0.000000 


0.000000 


0.000000 




949 


0.000000 


0.000000 


0.000000 




950 


0.000000 


0.000000 


0.000000 




951 


0.000000 


0.000000 


0.000000 




952 


0.876453 


0.876453 


0.000000 


0.00 


953 


0.000000 


0.000000 


0.000000 




954 


0.000000 


0.000000 


0.000000 




955 


0.000000 


0.000000 


0.000000 




956 


0.000000 


0.000000 


0.000000 




957 


0.000000 


0.000000 


0.000000 




958 


0.000000 


0.000000 


0.000000 




959 


0.000000 


0.000000 


0.000000 




960 


23.598419 


23.701996 


0.103577 


0.44 


961 


37.747559 


37.457550 


-0.290009 


-0.77 


962 


96.039429 


96.512787 


0.473358 


0.49 


963 


165.911163 


166.015625 


0.104462 


0.06 


964 


0.000000 


0.000000 


0.000000 




965 


0.000000 


0.000000 


0.000000 




966 


6202.445312 


6205.820312 


3.375000 


0.05 


967 


6308.570312 


6311.398438 


2.828125 


0.04 


968 


7345.796875 


7359.914062 


14.117188 


0.19 


969 


5594.445312 


5596.476562 


2.031250 


0.04 



312 



Final Report, July 1993 



970 


5591.109375 


5593.156250 


2.046875 


0.04 


971 


4557.953125 


4514.187500 


-43.765625 


-0.96 


973 


3360.766113 


3352.695312 


-8.070801 


-0.24 


973 


5838.195312 


5840.593750 


2.398438 


0.04 


974 


6087.726562 


6074.789062 


-12.937500 


-0.21 


975 


0.000000 


0.000000 


0.000000 




976 


5648.6S6250 


5635.320312 


-13.335958 


-0.24 


977 


3735.716797 


3735.937500 


0.220703 


0.01 


978 


392.843262 


392.895996 


0.052734 


0.01 


979 


4108.789062 


4113.570312 


4.781250 


0.12 


980 


2.784199 


2.784199 


O.TOSOOO 


0.00 


981 


5562.179688 


5559.757812 


-2.421875 


-0.04 


982 


273.306641 


259.376465 


-13.930T76 


-5.10 


983 


3403.514160 


3404.597168 


1. OB 3 008 


0.03 


984 


1.000000 


1.000000 


0.000000 


0.00 


985 


0.000000 


0.000000 


0.000000 




986 


0.000000 


0.000000 


0.000000 




987 


846.847656 


846.457520 


-0.390137 


-O.OS 


988 


3555.613281 


3434.521484 


-121.091797 


-3.41 


989 


97.254059 


96.060883 


-1.193176 


-1.23 


990 


453.313965 


453.246582 


-0.067353 


-0.01 


991 


171,714905 


171.821136 


0.106232 


0.06 


992 


207.586731 


207.833557 


0.246826 


0.12 


993 


277.473145 


277.477539 


0.004395 


0.00 


994 


0.000000 


0.000000 


0.000000 




995 


0.000000 


0.000000 


0.000000 




996 


0.229589 


0.231791 


0.002202 


0.96 


997 


0.000000 


0.000000 


0.000000 




998 


0.000000 


0.000000 


0.000000 




999 


1.690277 


1.879307 


0.189030 


11.18 


1000 


0.000000 


0.000000 


0.OT566 




1001 


0.000000 


0.000000 


0.000000 




1002 


0.000000 


0.000000 


0.000000 




1003 


0.000000 


0.000000 


0.000000 




1004 


0. 000000 


0.000000 


0.000000 




1005 


0.000000 


0.000000 


0.000000 




1006 


0.000000 


0.000000 


0.000000 




1007 


0.000000 


0.000000 


0.000000 




1008 


0.000000 


0.000000 


0.000000 




1009 


0.000000 


0.000000 


0.000000 




1010 


0.000000 


0.000000 


0.000000 




1011 


0.000000 


0.000000 


o .000000 




1012 


0.000000 


0.000000 


0.000000 




1013 


0.000000 


0.000000 


0.000000 




1014 


0.000000 


0.000000 


0.000000 




1015 


0.000000 


0.000000 


0.000000 




1016 


0. 000000 


0.000000 


0.000000 




1017 


211.113312 


236.889771 


25.776459 


12.21 


1018 


785.164062 


799.503906 


14.339844 


1.83 


1019 


88.783844 


99.807678 


11.023834 


12.42 


1020 


0. 000000 


0.000000 


0.000000 




1021 


0.000000 


0.000000 


0.000000 




1022 


0.000000 


0.000000 


0.000000 




1023 


0.000000 


0.000000 


0.000000 




1024 


0.000000 


0.000000 


0.000000 




1025 


0.000000 


0.000000 


0.000000 




1026 


3395.500977 


3396.104004 


0.603027 


0.02 


1027 


462.711914 


452.439453 


-10.272461 


-2.22 


1028 


79.743622 


79.680115 


-0.063507 


-0.08 


1029 


474.878418 


464.601074 


-10.277344 


-2.16 


1030 


16094.046875 


16038.218750 


-55.82*125 


-0.35 


1031 


0.000000 


0.000000 


0.000000 




1032 


131146.500000 


131863.750000 


717.250000 


0.55 


1033 


130100.125000 


130755.500000 


655.375000 


0.50 


1034 


132725.375000 


132462.875000 


-262.500000 


-0.20 


1035 


0.000000 


0.000000 


0.000000 




1036 


0. 000000 


0.000000 


0.000000 




1037 


0.000000 


0.000000 


o.dtfosoe 




1038 


-197.427338 


-236.117035 


-38 .619697 


19.60 


1039 


988.427246 


928.875977 


-59.551270 


-6.02 


1040 


0.000000 


0.000000 


0.000000 




1041 


1730.941895 


1729.177734 


-1.764160 


-0.10 


1042 


1727.729980 


1756.429688 


28.699707 


1.66 


1043 


1335.730957 


1373.146973 


37.41«¥l6 


2.80 


1044 


1349.046875 


1401.261230 


52.214355 


3.87 


1045 


0.000000 


0.000000 


0.000000 




1046 


95.691711 


95.923187 


0.231476 


0.24 


1047 


27.304840 


27.388367 


0,083527 


0.31 


1048 


165.844391 


165.928253 


0.083862 


0.05 


1049 


37.747559 


37.457550 


-0.290009 


-0.77 


1050 


26.894775 


28.839233 


-0.055542 


-0.19 


1051 


129382.375000 


129921.375000 


539.000000 


0.42 


1052 


128859.625000 


129207.125000 


347.500000 


0.27 


1053 


132011.875000 


131326.250000 


-685.625000 


-0.52 


1054 


0. 000000 


0.000000 


0.000000 




1055 


0.000000 


0.000000 


0.000300 




1056 


0.000000 


0.000000 


0.000000 




1057 


0. 000000 


0.000000 


0.000000 




1058 


: 0. 000000 


0.000000 


0.000000 




1059 


0. 000000 


0.000000 


0.000000 




1060 


: 0. 000000 


0.000000 


0.000000 




1061 


0. 000000 


0.000000 


0.066060 




1062 


0. 000000 


0.000000 


0.000000 




1063 


51.000000 


51.000000 


0.000000 


0.00 


1064 


51.000000 


51.000000 


0.000000 


0.00 


1065 


51.000000 


51.000000 


0.000000 


0.00 


1066 


: 15.000000 


15.000000 


0.000000 


0.00 


1067 


: 45.000000 


45.000000 


0.000000 


0.00 



Final Report, July 1993 



313 



1068 


: 15.000000 


15.000000 


0.000000 


0.00 


1069 


: 15.000000 


15.000000 


0.000000 


0.00 


1070 


: 45.000000 


45.000000 


0.000000 


0.00 


1071 


i 15.000000 


15.000000 


0.000000 


0.00 


1072 


: 30.500000 


30.500000 


0. 000000 


0.00 


1073 


: 36.000000 


36.000000 


0.000000 


0.00 


1074 


: 25.000000 


25.006666 


0.000000 


0.00 


1075 


: 25.000000 


25.000000 


0.000000 


0.00 


1076 


: 33.513794 


33.513794 


0.000000 


0.00 


1077 


: 33.513794 


33.513794 


0.000000 


0.00 


1078 


: 0.000000 


0.000000 


0.000000 




1079 


: 0.000000 


0.000000 


.000000 




1080 


: 0.000000 


0.000000 


0.000000 




1081 


: 0.000000 


0.000000 


0.000000 




1082 


0.000000 


0.000000 


0.000000 




1083 


: 0.000000 


0.000000 


0.000000 




1084 


: 0.000000 


0.000000 


0.000000 




1085 


: 0.000000 


0.000000 


0.000000 




1086 


: 0.000000 


0.000000 


0.000000 




1087 


: 0.000000 


0.000000 


0.000000 




1088 


: 0.000000 


0.000000 


0.000000 




1089 


0.000000 


0.000000 


0.000000 




1090 


0.000000 


0.000000 


0.000000 




1091 


0.000000 


0.000000 


0.000000 




1092 


0.000000 


0.000006 


0.000000 




1093 


0.000000 


0.000000 


0.000000 




1094 


0.000000 


0.000000 


0.000000 




1095 


0.000000 


0.000000 


0.000000 




1096 


0. 000000 


0.000000 


0.000000 




1097 


0.000000 


0.000000 


0. 000000 




1098 


0.000000 


0.000000 


0.000000 




1099 


0.000000 


0.000000 


0.000000 




1100 


0.000000 


0.000000 


0.000000 




1101 


395.000000 


395.000000 


0.000000 


0.00 


1102 


30.356171 


29.773682 


-0.582489 


-1.92 


1103 


27.270630 


27.931366 


0.660736 


2.42 


1104 


23.849548 


24.461273 


0.611725 


2.56 


1105 


27.682709 


27.383484 


-0.299225 


-1.08 


1106 


27.026703 


27.531891 


0.505188 


1.87 


1107 


1.202293 


1.250208 


0.047915 


3.99 


HOB 


0.981451 


0.982236 


0.000785 


0.08 


1109 


1.154089 


1 . 155670 


0.001581 


0.14 


1110 


1.137697 


1.159491 


0.021793 


1.92 


1111 


1.133795 


1.155737 


0.021942 


1.94 


1112 


1.040773 


1.032728 


-0.008045 


-0.77 


1113 


0.929834 


0.903496 


-0.026338 


-2.83 


1114 


0.933034 


0.906430 


-0.026604 


-2.85 


1115 


1.016426 


1.014395 


-0.002031 


-0.20 


1116 


0.962189 


0.938432 


-0.023757 


-2.47 


1117 


1.181183 


1.202496 


0.021313 


1.80 


1118 


1.004492 


1.022034 


0,017542 


1.75 


1119 ' 


922.400391 


922.549805 


0.149414 


0.02 


1120 


5853.007812 


5855.554688 


2.546875 


0.04 


1121 


921.980957 


922.193359 


0.212402 


0.02 


1122 * 


7707.125000 


7707.125000 


0.000000 


0.00 


1123 


1.250298 


1.250298 


0.000000 


0.00 


L124 


153.033630 


153.397064 


0.363434 


0.24 


L125 


15744.898438 


15733.726562 


-11.171875 


-0.07 


1126 . 


153.457642 


153.784271 


0.326630 


0.21 


1127 


31888.382812 


31888.382812 


0.000000 


0.00 


1128 : 


1.204300 


1.204300 


0.000000 


0.00 


1129 : 


0.000000 


0.000000 


0.000000 




1130 : 


0.372473 


0.372921 


0.000447 


0.12 


1131 


0.673990 


0.674193 


0.000202 


0.03 


1132 


0.961010 


0.958860 


-0.002150 


-0.22 


1133 . 


0.984367 


0.973787 


-0.010580 


-1.07 


1134 


1.000000 


1.000000 


0.000000 


0.00 


1135 . 


0.876453 


0.876453 


0.000000 


0.00 


1136 . 


0.877076 


0.874804 


-0.002273 


-0.26 


1137 


2.917759 


2.917759 


0.000000 


0.00 


1138 . 


2.920059 


2.919474 


-0.000586 


-0.02 


1139 : 


0.000000 


0.000000 


0.000000 




1140 : 


1.000000 


1.000000 


0.000000 


0.00 


1141 : 


1.000000 


1.000000 


0.000000 


0.00 


1142 : 


0.000000 


0.000000 


0. 000000 




1143 : 


75.010498 


131.967773 


56.957275 


75.93 


1144 : 


84.308472 


97.011780 


12.703308 


15.07 


1145 : 


1.494102 


2.326550 


0.832447 


55.72 


1146 : 


-0.334679 


-0.521147 


-0.186468 


55.72 


1147 : 


265.512695 


286.891602 


21.378906 


8.05 


1148 : 


208.515900 


233.599915 


25.014015 


12.03 


1149 : 


-14.747295 


-21.245453 


-6.498158 


44.06 


1150 : 


-11.091982 


-15.979502 


-4.887520 


44.06 


1151 : 


1.242491 


1.453331 


0.210840 


16.97 


1152 : 


-0.264773 


-0.309703 


-0.044930 


16.97 


1153 : 


1552.496582 


1568.353516 


15.856934 


1.02 


1154 : 


1754.432129 


1752.087891 


-2.344238 


-0.13 


1155 : 


-538.620117 


-582.543457 


-43.923340 


8.15 


1156 ; 


-322.135254 


-339.190918 


-17.055664 


5.29 


1157 : 


66.000000 


66.000000 


0. 000000 


0,00 


1158 : 


30.399994 


30.399994 


0.000000 


0.00 


1159 : 


0.000000 


0.000000 


0.000000 




1160 ; 


0.000000 


0.000000 


0.000000 




1161 : 


0.000000 


0.000000 


0.000000 




1162 : 


0.000000 


0.000000 


0.000000 




1163 : 


0.000000 


0.000000 


0.000000 




1164 : 


0.000000 


0.000000 


0.000000 




U« : 


0.000000 


0.000000 


0.000000 





314 



Final Report, July 1993 



1166 : 


0. 000000 


0.000000 


0.000000 




1167 : 


32.417328 


34.355530 


1.938202 


5.98 


1168 : 


-1764.123047 


-1942.351074 


-178.228027 


10.10 


1169 : 


-1240.480957 


-1548.384766 


-307.903809 


24.82 


1170 : 


-713,485840 


-1136.531738 


-423.645B98 


59.29 


1171 : 


0.000000 


0.000000 


0.000000 




1172 : 


27.403717 


27.555206 


0.151489 


0.55 


1173 : 


541.725098 


566.392578 


24.667480 


4.55 


1174 : 


23.792389 


25.756317 


1.963928 


8.25 


1175 : 


492.237793 


493.498535 


1.260742 


0.26 


1176 : 


18.034485 


18.133728 


0.099243 


0.55 


1177 : 


447.876953 


468.583496 


20.706543 


4.62 


1178 : 


16.324646 


17.683960 


1.359314 


8.33 


1179 : 


0.000000 


0.000000 


0.000000 




1180 : 


33.235779 


34.780548 


1.544769 


4.65 


1181 : 


14513.359375 


14511.671875 


-1.687500 


-0.01 


1182 : 


12774624.000000 


12775666.000000 


1047.000000 


0.01 


1183 : 


566197.250000 


571219.875000 


5022.625000 


0.89 


1184 : 


201090.125000 


215010.625000 


13926.500000 


6.92 


1185 : 


236.782410 


237.057434 


0.275024 


0.12 


1186 : 


246.296967 


246.676880 


0.379913 


0.15 


1187 : 


379.218750 


377.831055 


-1.387695 


-0.37 


1188 : 


38.898804 


38.714111 


-0.1846*2 


-0.47 


1189 : 


450.126465 


458.618652 


6.492188 


1.89 


1190 : 


55.162903 


57.326813 


2.163910 


3.92 


1191 : 


607.890625 


609.462891 


1.572266 


0.26 


1192 : 


0.000000 


0.000000 


0.000000 




1193 : 


0.000000 


0.000000 


0.000000 




1194 : 


0.000000 


0.000000 


0.000000 




1195 : 


0.000000 


0.000000 


0.000000 




1196 : 


0.000000 


0.000000 


0.000000 




1197 : 


0.000000 


0.000000 


0.000000 




1198 : 


0.000000 


0.000000 


0.000000 




1199 : 


-1.054302 


-1.053318 


0.000984 


-O.09 


1200 ; 


0.000000 


0.000000 


0.000000 




1201 : 


0.000000 


0.000000 


0.000000 




1202 : 


0.000000 


0.000000 


0.000000 




1203 : 


0.000000 


0.000000 


0.000000 




1204 : 


0.000000 


0.000000 


0.060000 




1205 : 


0.000000 


0.000000 


0.000000 




1206 : 


0.000000 


0.000000 


0. 000000 




1207 : 


0.000000 


0.000000 


0.000000 




1208 : 


0.000000 


0.000000 


0.000000 




1209 : 


0.000000 


0.000000 


0.000000 




1210 : 


0.000000 


0.000000 


0.000000 




1211 : 


0.000000 


0.000000 


0.000000 




1212 : 


0.000000 


0.000000 


0.000000 




1213 : 


0.000000 


0.000000 


0.000000 




1214 : 


0.000000 


0.000000 


0.000000 




1215 : 


0. 000000 


0.000000 


0.000000 




1216 : 


0.000000 


0.000000 


0.000000 




1217 : 


0.000000 


0.000000 


0. 000000 




1218 : 


0.000000 


0.000000 


0.060000 




1219 : 


0.000000 


0.000000 


0.000000 




1220 : 


0.000000 


0. 000000 


0.006000 




1221 : 


0.000000 


0.000000 


0.066000 




1222 : 


0.000000 


0.000000 


0.000060 




1223 : 


0. 000000 


0.000000 


0.006660 




1224 : 


0.000000 


0.000000 


0.000000 




1225 : 


0.000000 


0.000000 


0.000000 




1226 : 


0.000000 


0.000000 


0.000000 




1227 : 


0.000000 


0.000000 


0.000000 




1228 : 


0.000000 


0.000000 


0.000000 




1229 : 


0.000000 


0.000000 


0.666000 




1230 : 


0.000000 


0.000000 


0.000000 




1231 : 


0.000000 


0.000000 


0.006000 




1232 : 


0.000000 


0.000000 


0.666600 




1233 : 


0.000000 


0.000000 


0.066600 




1234 : 


0.000000 


0.000000 


6.666606 




1235 : 


0. 000000 


0.000000 


0.000660 




1236 : 


0.000000 


0.000000 


0.006660 




1237 : 


0.000000 


0.000000 


0.000000 




1238 : 


0.000000 


0.000000 


0.000000 




1239 : 


0.000000 


0.000000 


0.000000 




1240 : 


0.000000 


0.000000 


0.000660 




1241 : 


0.000000 


0. 000000 


0.000600 




1242 : 


0.000000 


0.000000 


0.666600 




1243 : 


0.000000 


0.000000 


0.666666 




1244 : 


0.000000 


0.000000 


0.000006 




1245 : 


0.000000 


0.000000 


0.000000 




1246 : 


0.000000 


0.000000 


0.000000 




1247 : 


0.000000 


0.000000 


0.000006 




1248 : 


0.000000 


0.000000 


6.066666 




1249 : 


0.000000 


0.000000 


0.600060 




1250 : 


0.697200 


0.697200 


0,000000 


0.00 


1251 : 


-1.181999 


-1.181999 


6.066000 


-0.00 


1252 : 


0.520800 


0.520800 


0.000000 


0.00 


1253 : 


0.000000 


0.000000 


0.000066 




1254 : 


0.050000 


0.050000 


0.000060 


0.06 


1255 : 


0.000000 


0.000000 


0.660600 




1256 : 


0.000000 


0.000000 


0.000000 




1257 ; 


0.000000 


0.000000 


0.066000 




1258 : 


0.000000 


0.000000 


0.000066 




1259 : 


0.000000 


0.000000 


0.000066 




1260 : 


0.000000 


0.000000 


0. 000660 




1261 : 


0.000000 


0.000000 


0.000600 




1262 : 


0.000000 


0.000000 


0.066600 




1263 : 


0.000000 


0.000000 


0.066066 





Final Report, July 1993 



315 



1264 : 


0.000000 


0.000000 


0.000000 




1265 : 


0.000000 


0.000000 


0.000000 




1266 : 


0.000000 


0.000000 


0.000000 




1261 : 


0.000000 


0.000000 


0.000000 




1268 : 


0.000000 


0.000000 


O.OOOOOO 




1269 : 


0.000000 


0.000000 


0.000000 




1270 : 


0.000000 


0.000000 


0.000000 




1271 : 


0.000000 


0.000000 


0.000000 




1272 : 


0.000000 


0.000000 


0.000000 




1273 : 


0.000000 


0.000000 


0.000000 




1274 : 


0.000000 


0.000000 


0.000000 




1275 : 


0.000000 


0.000000 


0.000000 




1276 : 


0.000000 


0.000000 


0.000000 




1277 : 


0.000000 


0.000000 


0.000000 




1278 : 


0.000000 


0.000000 


0.000000 




1279 : 


0.000000 


0.000000 


0.000000 




1290 : 


0.000000 


0.000000 


0.000000 




12S1 : 


0.000000 


0.000000 


0.000000 




1282 : 


0.000000 


0.000000 


0.000000 




1283 : 


0.000000 


0.000000 


0.000000 




1284 : 


0.000000 


0.000000 


0.000000 




1285 : 


0.000000 


0.000000 


0.000000 




1286 : 


0.000000 


0.000000 


0.000000 




1287 : 


0.000000 


. oooooo 


0.000000 




1288 : 


0.000000 


0.000000 


0.000000 




1289 : 


0.000000 


0.000000 


0.000000 




1290 : 


0.000000 


0.000000 


0.000000 




1291 : 


1.000000 


1.000000 


0.000000 


0.00 


1292 : 


1.000000 


1.000000 


0.000000 


0.00 


1293 : 


1.007999 


1.007999 


0.000000 


0.00 


1294 : 


1.000000 


1.000000 


0.000000 


0.00 


1295 : 


1.002886 


1.002886 


0.000000 


0.00 


1296 : 


1.000000 


1.000000 


0.000000 


0.00 


1297 : 


0.000000 


0.000000 


0.000000 




1298 : 


0.000000 


0.000000 


0.000000 




1299 s 


0.000000 


0.000000 


0.000000 




1300 : 


0.000000 


0.000000 


O.OOOOOO 




1301 : 


o.oooooo 


0.000000 


0.000000 




1302 : 


0.016555 


0.016555 


0.000000 


0.00 


1303 


-0.030621 


-0.030621 


0.000000 


-0.00 


1304 


1.000000 


1. oooooo 


0.000000 


0.00 


1305 


1.246698 


1.246698 


0.000000 


0.00 


1306 


1.204300 


1.204300 


0.000000 


0.00 


1307 


0.000000 


0.000000 


0.000000 




1308 


1.000000 


1.000000 


0.000000 


0.00 


1309 


1.000000 


1.000000 


0.000000 


0.00 


1310 


0.000000 


0.000000 


0.000000 




1311 


0.000000 


0.000000 


O.OOOOOO 




1312 


0.000000 


0.000000 


0.000000 




1313 


0.000000 


0.000000 


0.000000 




1314 


0.000000 


0.000000 


0.000000 




1315 


0.000000 


0.000000 


0.000000 




1316 


0,000000 


0.000000 


0.000000 




1317 


: 0.000000 


0.000000 


0.000000 




1318 


: 1,000000 


1.000000 


0.000000 


0.00 


1319 


: 0.000000 


0.000000 


0.000080 




1320 


: 0.000000 


0.000000 


0.000000 




1321 


: 0.000000 


0.000000 


0.000000 




1322 


: 0.000000 


0.000000 


0.000000 




1323 


: 0.000000 


0.000000 


0.000000 




1324 


: 778.259766 


778.259766 


0.000000 


0.00 


1325 


: 32.173981 


32.173981 


0.000000 


0.00 


1326 


3.141588 


3.141588 


0.000000 


0.00 


1327 


: 2.015999 


2.015999 


0.000000 


0.00 


1328 


: 1545.429688 


1545. 429688 


0.000000 


0.00 


1329 


: 144.000000 


144.000000 


0.000000 


0.00 


1330 


: O.OOOOOO 


0.000000 


0.000000 




1331 


; 0.000000 


0.000000 


0.000660 




1332 


: 0.000000 


0.000000 


0.000000 




1333 


: 0.000000 


0.000000 


0.000000 




1334 


: 0.000000 


0.000000 


0.000000 




1335 


: 0.000000 


0.000000 


0.000000 




1336 


: O.OOOOOO 


O.OOOOOO 


6'. OOOOOO' 




1337 


: 0.000000 


0.000000 


0.000000 




1338 


: 0.000000 


0.000000 


0.000000 




1339 


: 0.000000 


0.000000 


0.000000 




1340 


: 0.000000 


0.000000 


0.000000 




1341 


: 0.000000 


0.000000 


0.000000 




1342 


: 0.000000 


0.000000 


0.066066 




1343 


: 0.000000 


0.000000 


0.000000 




1344 


: 0.000000 


0.000000 


0.000000 




1345 


: 0.000000 


0.000000 


0.000000 




1346 


: 0.000000 


0.000000 


0.000000 




1347 


: 0.000000 


0.000000 


0.000000 




1348 


: 0.000000 


0.000000 


0.000000 




1349 


: 0.000000 


0.000000 


0.006000 




1350 


: 10402.500000 


10402.500000 


0.000600 


0.00 



316 



Final Report, July 1993 



Standard EDIS Execution Transcript 



A.5.3MCC Leak Example: Standard EDIS execution transcript 

Standard EDIS Execution Transcript 

NEXPEKT Serial Number 1-2.0B-S4X1-051091-1458 
Copy for Client of Neuron Data, Inc. 

# . NEXPEKT. - Copyright (C) 1986 - 1990 by NEURON DATA. Copyright is claimed in both the underlying comput- 
er program and the resulting output in the form of an audiovisual work. 

# Customer or User is not permitted to make any copies of this software (NEXPEKT) for any purpose. This software 
is a confidential trade secret of NEURON DATA Inc. Refer to the license agreement. 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = FUEL_FLOW_CTRL_2 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO^ 

# Next behavior to expand 

# CONTROL JDBreCT2.is_now = F101_4 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_12 

# Ignoring pin deviation at HIPUMP 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = HPFP_3 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIOJL3 

# One fault in behavior 

# CONTROL_OBJECT3.best_faulty_behavior = HPFP_7 

# Fault type 

# CONTROL_OBJECT3.best_currentJault = LOWJEFFICIENCY 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = HPFT_6 

# No viable expansion! 

# New best scenario 

# CONTROL JDBJECT.current_scenario = SCENARIO_14 

# Next behavior to expand 

# CONTROL JDBJECT2.is_now = NOZZLEl^S 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_15 

# No fault! 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = MCC_7 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_20 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = MOVJ7 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_28 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = O204_5 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_30 

# Next behavior to expand 

# COOTROL_OBJECT2.is_now = M104J7 

# New best scenario 

# CONTROL_OBJECTxurrent_scenario = SCENARIO_43 

# Next behavior to expand 

# CONTROL OBJECT2.is now = O205 5 



Final Report, July 1993 317 



# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_49 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = M101_8 

# New best scenario 

# CONTROL JDBJECT.current_scenario = SCENARIO_52 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = OFOV_6 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_54 

# Next behavior to expand 

# CbNTO6L_OBJECT2.is_now = OPB_7 

# New best scenario 

# CONTROL J)BJECTxurrent_scenario = SCENARIO_60 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = HP0T_9 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_61 

# Next behavior to expand 

# CONTR0L_OBJECT2.is_now = HPOP_PBP_8 

# New best scenario 

# CONTROL jOBJECT.current_scenario = SCENARIO_63 

# One fault in behavior 

# CONTROL J)BJECT3.bestJaultyJ>ehavior = HFOP_PBP_14 

# Fault type 

# CONTROL jDBJECT3.best_current_fauIt = LOW_EFFICIENCY 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = O201_4 

# New best scenario 

# CONTROL OBJECTcurrent scenario = SCENARIO_64 

# No fault! 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = O201 JO 

# New best scenario 

# CONTROL_OBJECT.cunentj>cenario = SCENARIO_72 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPOP_6 

# New best scenario 

# CONTROL^BJECTxurrent^scenario = SCENARIO_73 

# One fault in behavior 

# CONTROL_OBJECT3.best Jaulty^behavior = LPOP _10 

# Fault type 

# CONTROL JDBJECn.best_current_fault = LOW_EFFICIENCY 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPOT_5 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_87 

# Next behavior to expand 

# CONTROLJDBJECT2.is_now =: O203_7 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_89 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = O203_10 

# New best scenario 



328 Final Report, July 1993 



# CONTROL_OBJECT.currcnt_scenario = SCENARIO_94 

# Next behavior to expand 

# CONTROL_OBJECT2as_now = HPFT_10 

# New best scenario 

# CONTROL JDBJECT.current_scenario = SCENARIO_95 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = HGM_9 

# New best scenario 

# CONTROL_OBJECTxunent_scenario = SCENARIOJL10 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = DIFFUSER_7 

# New best scenario 

# CONTROL_OBJECT.cunent_scenario = SCENARIO_117 

# Next behavior to expand 

# CONTROL jDBJECT2.is_now = CCVJ5 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_12I 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = MKER_10 

# New best scenario 

# CONTROL jDBJECT.current_scenario = SCENARIO_134 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = NZL_COOL_10 

# New best scenario 

# CONTROL_OBJECTxurrent_scenario = SCENARIO_135 

# Next behavior to expand 

# CONTROL JDBJECT2.is_now = MFV_7 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIOJL36 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPFP J 

# New best scenario 

# CONTROLJ>BJECT.current_scenario = SCENARIO_147 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPFT _3 

# New best scenario 

# CONTROL JDBJECT.current_scenario = SCENARIO_156 

# Next behavior to expand 

# CONTROL JDBJECT2.is_now = F109_6 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_170 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = MCC_COOL_9 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_150 

# Fault assumption looks wrong: increase estimated cost 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIOJ74 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPOT_78 

# New best scenario 

# CONTROL_OBJECT.cunent_scenario = SCENARIO_198 

# No fault! 

# Next behavior to expand 



m 



Final Report, July 1993 r 319 



# CONTROL_OBJECT2 is _now = O203J3 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_199 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = HPFTJO 

# New best scenario 

# CONTROL JDBJECTxurrent_scenario = SCENARIO_200 

# Next behavior to expand 

# CONTROL.OBJECT2.is_.now = HGM_23 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_215 

# Next behavior to expand 

# CONTCOL J)BJECT2.is_now = DIFFUSERJ7 

# New best scenario 

# CONTROLjOBJECT.oinent^scenario = SCENARIO_222 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = CCVJ1 

# New best scenario 

# CONTROL JDBJECTxurrent_scenario = SCENARIO_226 

# Next behavior to expand 

# CONTROLJDBJECT2.is_now = MKER_44 

# New best scenario 

# CONTROL JDBJECTxurrent_scenario = SCENARIO_239 

# Next behavior to expand 

# CONTROL jDBJECT2.is_now = NZL_COOL_17 

# New best scenario 

# CONTROL jDBJECTxurrent_scenario = SCENARIO_240 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = MFV_21 

# New best scenario 

# CONTROL jDBJECTxurrent_scenario = SCENARIO_241 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPFP_3 

# New best scenario 

# CONTROL_OBJECTxurrent_scenario = SCENARIO_254 

# One fault in behavior 

# C0NTR0L_0BJECT3.best _faulty_behavior = LPFPJ7 

# Fault type 

# CONTROL_OBJECT3.best_current _fault = LOW_EFFTCIENCY 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPFT_24 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_258 

# Next behavior to expand 

# CONTROL J)BJECTO.is_now = F109_17 

# New best scenario 

# CONTROL JDBJECTxurrentjicenario = SCENARIO_266 

# Next behavior to expand 

# CONTROL jDBJECT2.is_now = MCC_COOL_19 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_253 

# One fault in behavior 

# CONTROL_OBJECT3.best_faulty_behavior = LPFP_16 

# Fault type 



320 , Final Report, July 1993 



# CONTROL J3BJECT3.best_current_fault = LOW_EFnCIENCY 

# Fault assumption looks wrong: increase estimated cost 

# New best scenario 

# CONTROL J3BJECT.current_scenario = SCENARIO_253 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPFT_36 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_269 

# Next behavior to expand 

# CONTROL^OBJECm.is^ow = F109 J27 

# New best scenario 

# CONTROL_OBJECT.curTent_scenario = SCENARIO_277 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = MCCCOOL_28 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_270 

# Fault assumption looks wrong: increase estimated cost 

# New best scenario 

# CONTROL J3BJECT.current_scenario = SCENARIO_252 

# Next behavior to expand 

# CONTROL jDBJECT2.is_now = LPFT_45 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_286 

# No fault! 

# Next behavior to expand 

# CONTROL J3BJECT2.is_now = F109_37 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_300 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = MCC_COOL_37 

# New best scenario 

# CONTROL JDBJECT.current_scenario = SCENARIO_301 

# One fault in behavior 

# CONTROL jDBJECT3.bestJaulty_behavior = MCC_COOL_43 

# Fault type 

# CONTROL J3BJECn.best_current_fault = LEAK 

# Next behavior to expand 

# CONTROL JDBJECT2.is_now = FPB_6 

# New best scenario 

# CONTROL_OBJECTxurrent_scenario = SCENARIO_310 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = FPOV Jl 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO = 318 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = O206_7 

# New best scenario 

# CONTROL J3BreCT.currentjscenario = SCENARIO_319 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = F107_10 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_325 

# Next behavior to expand 

# CONTROL OBJECT2.is now = M103 14 



Final Report, July 1993 321 



# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_328 

# Next behavior to expand 

# CONTROLjOBJECT2.is_now = F110_9 

# New best scenario 

# CONTROL J3BJECTxurrent_scenario = SCENARIO_329 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = F108_8 

# New best scenario 

# CONTROL OBJECT.current scenario = SCENARIO 330 



322 Final Report, July 1993 



PBM Qualitative Parameter Values 



A.5.4MCC Leak Example: PBM qualitative data file 

PBM Qualitative Parameter Values 

\MCC_PBM_tcmpIate.MR\="NORMAL" 

\JLPFP_PBM_template.pin\="NORMAL" 

\LPOPJPBMltemplate.pin\="NORMAL" 

\LPFP_PBM_template.Tin\="NORMAL" 

\LPOP_PBM_template.Tin\="NORMAL" 

\MOV_PBM_template.p_diff\="NORMAL" 

\O201 PBM_tcmplate.p_dim="NORMAL" 

\LPCf PBM_tempIate.p_diff\="NORMAL" 

VMFV_PBM_tcmplatc.p_difFv="HIGH" 

\MOV_PBM_template.p_dim="NORMAL" 

\OPOV_PBM_tempIate.p_diff\="LOW" 

\FPOV_PBM_template.p_diff\="L0W" 

\F110_PBM_templatc.p_diff\="NORMAL" 

\CCV_PBM_template.p_diff\="NOTUvlAL" 

\NZL_COOL_PBM_tempIate.T_di£E\="NORMAL" 

\MCC_COOL_PBM_template.T_diff\="L0W" 

\LPFP_PBM_tcmpIate.p_diff\="LOW" 

\HPFP_PBM_template.p_dim="NORMAL" 

\LPOP_PBM_template.p_diff\="NORMAL" 

\HPOP~PBP_PBM tcmpIate.p_diff\="NORMAL" 

\LPFP_PBM_tempiatc.MechPWR\="LOW" 

\HPFP_PBM_template.McchPWR\="NORMAL" 

\LPFT_PBM_template.MechPWR\="LOW" 

\HPFT_PBM_template.MechPWR\=''NORMAL" 

\LPOP_PBMjemplate.MechPWR\="NORMAL" 

\HPOP_PBP_PBM_template.MechPWR\="NORMAL" 

\LPOT_PBM_template.MechPWR\= , 'NORMAL" 

\HPOT_PBM_template.MechPWR\= ,, NORMAL" 

\MKER_PBM_tcmpIatc.pin\="NORMAL" 

\F107_PBM_template.pin\="NORMAL" 

\FPB_PBM_tcmplatc.poutVNORMAL" 

\HPFT_PBM_tcmplatc.pin\= ,, NORMAL" 

\HGMJ > BM_template.pin\=' , NORMAL" 

\OPB_PBM_template.pout\= M NORMAL" 

\MFV_PBM_template.pin\= M NORMAL" 

\F101_PBM_tempIate.pout\="LOW" 

\LPFT_PBM_template.pin\="NORMAL" 

\HPFT_PBM_template.pin\="NORMAL" 

\HPOP_PBP_PBMjemplatc.pin\="NORMAL" 

\O201_PBM_template.pout\=*'NORMAL" 

\LPOT_PBM_template.pin\='*NORMAL" 

\M104 PBM_template.pin\="NORMAL" 

\HPOT_PBM_template.pin\= ,, NORMAL" 

\OPOV_PBM_template.pin\="NORMAL" 

\M101_PBM_template.pin\= B NORMAL M 

\OPB_PBM_template.pin_OX\="N0RMAL" 

\FPB_PBM_tcmplatc.pin_OX\="NORMAL" 

\LPFP_PBM_template.pout\="LOW" 

\F101_PBM_template.pin\="LOW" 

\HPFP_PBM_tempIate.pout\="NORMAL" 

\LPOP_PBM_template.pout\="NORMAL" 



Final Report, July 1993 323 



\O201_PBM_template.pin\="NORMAL" 

\HP0P_PBP_PBM_template.pout\="NORMAL" 

\MOV_PBM_template.pin\="NORMAL" 

\O204 PBM_template.pout\="NORMAL" 

\LPFT~PBM_template.pout\="NORMAL" 

\HPFT_PBM_template.pout\="NORMAL" 

\HPOT_PBM_template.pout\="NORMAL" 

\CCV_PBM_template.pin\="NORMAL" 

\MCC_COOL_PBM_template.pin\="NORMAL" 

\OPOV PBM_template.pout\="NORMAL" 

\NZL_ODOL_PBM_tcmplate.pin\="NORMAL" 

\FPOV PEMjemplate-poutV^NORMAL" 

\CCV PEMjemplatcpout^TNORMAL" 

VFPB^BMjemplatc.pin^'NORMAL" 

\F110 PBM_template.pout\="NORMAL" 

\OPB~PBM_template.pm\="NORMAL" 

\F108~PBM_tempIate.pout\="NORMAL" 

\MOV_PBM_template.pout\="NORMAL" 

\MCC_PBM_template.pin_OX\="NORMAL" 

\FPB_PBM_tcmpIate.MR\="NORMAL" 

\OPB_PBM_templatc.MR\= M fflGH" 

\MDOER_PBM template.Tout\="NORMAL" 

\F107_PBM_template.Tin\="NORMAL" 

\CCV_PBM_tcmplate.Tout\="NORMAL" 

\MKER_PBM_template.Tin\='*NORMAL" 

\NZL_COOL_PBM_template.Tout\="NORMAL" 

\MIXER_PBM_temp]ate.TinB\="NORMAL" 

\HPOP_PBP_PBM_templatc.Tin\= "NORMAL" 

\O201_PBM_tcmplatc.Tout\= ,, N0RMAL" 

\LPOT_PBM_temp]ate.Tq\="NORMAL" 

\LPOP_PBM_template.Tq\= w NbRMAL" 

\FPB_PBM_template.Tout\='^ORMAL" 

\HPFT_PBM_temp1ate.Tin\="NORMAL" 

\OPB_PBM_templatc.Tout\="HIGH" 

\HPOT_PBM_template.Tin\="HIGH" 

\HPFP_PBM_tcmplatc.Tin\="NORMAL" 

\F101_PBM_template.Tout\="NORMAL" 

\LPPT PBM_templatc.Tin\=' , NORMAL" 

\LPFP~PBM_templatc.Tout\="NORMAL" 

\F101J > BMjemplate.Tin\="NORMAL" 

\HPFTP_PBMjemplateTout\="NORMAL" 

\MFV_PBM_tempIate.Tin\="NORMAL" 

\LPFT_PBM_templatc.Tout\="NORMAL" 

\HPFT_PBM_template.Tout\="NORMAL" 

\HGM_PBM_tcmpTate.fin\="NORMAL" 

\LPOP_PBM_template.Tout\="NORMAL" 

\O201jmiJempIate:rin\=T*ORMAL'' 

\HPOP_PBP_PBMjcmpiatc.tout\= ,, NORMAL" 

\M104_PBM_tcmplate.Tin\="NORMAL" 

\LPOT_PBM_template.Tout\="NORMAL" 

\HPOT_PBM_template.Tout\="HIGH" 

\HGM_PBM_templatc.TinB\="HIGH" 

\MFV PBM_tempIate.Tout\="NORMAL" 

\DIFFUSER PBM_tcmplatc.Tin\="NORMAL" 



• 



324 



Final Report, July 1993 



\FPOV_PBM_templatc.Tout\="NORMAL" 

\FPB_PBM_temp1ate.Tm_0X\= M N0RMAL" 

\OPOV_PBM_tcraplate.Tout\= "NORMAL" 

\OPBJ > BMjemplateTin_OX\="NORMAL" 

\MFV~PBM_templatc.Vbar\="NORMAL" 

\DffFUSERPBM tcmplate.Vin\="NORMAL" 

\FPB_PBM_tempirte.Vin\="NORMAL" 

\F110_PBM_template.Vout\="NORMAL" 

\FPB PBM_template.Vin_OX\="NORMAL" 

\FPOV_PBM_tcmplatc.Vout\="NORMAL" 

\LPFP_PBM_tempIate.Vbar\="NORMAL" 

\HPFP PBM_tcmplatc.Vbar\="NORMAL" 

\HGM~PBM_tempIate.Vin\= w NORMAL" 

\HPFT~PBM_tempIate.Vout\="NORMAL" 

\LPFT_PBM_tcmpIate.Vin\="NORMAL" 

\F109_PBM_tcmplate.Vout\="NORMAL" 

\HPFT_PBM template.Vin\="NORMAL" 

\FPB_PBM_template.Vout\= ,, NORMAL" 

\MOV_PBM_tcmplate.Vbar\="NORMAL" 

\O204_PBM_tempIate.Vout\="NORMAL" 

\MCCIPBM_tempIate.Vm_OX\="NORMAL" 

\OPB_PBM_tempIate.Vin\="NORMAL" 

\F108_PBM_template.Vout\="NORMAL" 

\OPB PBM_tempIate.Vin_OX\="fflGH" 

\OPOV_PBM_template.Vout\="fflGH" 

VLPOP_PBM_template.Vbar\="NORMAL" 

\O201 PBM tempIate.Vin\="NORMAL" 

\O20rPBMjemplate.Vbar\="NORMAL" 

\HPOf.PBP_PBM_temp]ate.Vbar\="NORMAL" 

\O201 PBMjemplate.Vout\="NORMAL" 

\M104_PBM_template.Vin\="NORMAL" 

VLPOT_PBMjemplate.Vbar\="NORMAL" 

\O203_PBM_template.Vout\="NORMAL" 

\HPOT_PBM_tcmplatc.Vin\= M fflGH" 

VOPB.PBMjemplate.Vout^HIGH" 

\NZL_COOL_PBM template .Vbar\="NORMAL" 

\OPOV PBM_temprate.Vbar\="fflGH" 

\M101J > BM_temp]ate.Vout\="HIGH ,, 

\FPOV_PBM_template.Vbar\="NORMAL" 

\O206_PBM_template.Vout\= M NORMAL" 

\MIXER_PBM_template.ViiiB\="NORMAL" 

\NZL_COOL PBM template .Vout\="NORMAL" 

\MKER_PBM_template.Vout\="NORMAL" 

\F107_PBM_template.Vin\="NORMAL" 

\OPB PBM template.Vout\="fflGH" 

\HPOT_PBM_template.Viii\="HIGH" 

\MCC_COOL_PBM_template.Vbar\="HIGH" 



********** 



g Final Report, July 1993 325 



Execution Transcript: Using PBM 



A.5.5MCC Leak Example: Execution transcript using PBM 

Execution Transcript: Using PBM 

NEXPEKT Serial Number 1-2.0B-S4X1-051091-1458 
Copy for Client of Neuron Data, Inc. 

# . NEXPEKT. - Copyright (C) 1986 - 1990 by NEURON DATA. Copyright is claimed in both the underlying comput- 
er program and the resulting output in the form of an audiovisual work. 

# Customer or User is not permitted to make any copies of this software (NEXPEKT) for any purpose. This software 
is a confidential trade secret of NEURON DATA Inc. Refer to the license agreement. 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = FUEL_FLOW_CTRL_2 

# New best scenario 

# CONTROL JDBJECT.currentjranario = SCENARIO^ 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = F101_4 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_12 

# Ignoring pin deviation at HIPUMP 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = HPFP_3 

# New best scenario 

# CONTROL_OBJECTxurrent_scenario = SCENARIO_13 

# One fault in behavior 

# CONTROL_OBJECT3.best Jaultyjjehavior = HPFP_7 

# Fault type 

# CONTROL_OBJECT3.best_currentJault = LOWJEFFICIENCY 

# Next behavior to expand 

# CONTROL JDBJECT2.is_now = HPFT_6 

# No viable expansion! 

# New best scenario 

# CONTROL J3BJECTxurrent_scenario = SCENARIO_14 

# Next behavior to expand 

# CONTROL jDBJECT2.is_now = NOZZLEl_5 

# New best scenario 

# CONTROL_OBJECT.cuiTent_scenario = SCENARIO_15 

# No fault! 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = MCC_7 

# New best scenario 

# CONTROL_OBJECTxurrent_scenario = SCENARIO_20 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = MOV_7 

# New best scenario 

# CONTROL_OBJECTxurrent_scenario = SCENARIO_28 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = O204_5 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_30 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = M104J7 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_43 

# Next behavior to expand 

# CONTROL OBJECT2.is now = O205 5 



326 Final Report, July 1993 



# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_49 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = M101_8 

# New best scenario 

# CONTROL JDBJECT.current_scenario = SCENARIO_52 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = OPOV_6 

# New best scenario 

# CONTROL JDBJECT.current_scenario = SCENARIO_54 

# Next behavior to expand 

# CONTROL JDBJECT2.is_now = OPBJ7 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_60 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = HPOT_9 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO J51 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = HPOPJ>BP_8 

# New best scenario 

# CONTROL jDBJECT.current_scenario = SCENARIO_63 

# One fault in behavior 

# CONTROL_OBJECT3.best_faulty_behavior = HPOP_PBP_14 

# Fault type 

# CONTROL_OBJECT3.best_current_fault = LOW_EEFICIENCY 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = O201_4 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_64 

# No fault! 

# Next behavior to expand 

# CONTROL.OBJECT2.is.now = O201 JO 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIOJ72 

# Next behavior to expand 

# CONTROLjDBJECT2.is_now = LP0P_6 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO J73 

# One fault in behavior 

# CONTROL_OBJECT3.best_faulty_behavior = LPOP_10 

# Fault type 

# CONTROL.OBJECT3.best.current_fault = LO\VEFFTCIENCY 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LP0T_5 

# New best scenario 

# CONTROL JDBJECT.current_scenario = SCENARIO_89 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = O203J7 

# New best scenario 

# CONTROL jDBJECT.currentjjcenario = SCENARIO_93 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = HPFT_11 

# New best scenario 



Final Report, July 1993 327 



# CONTROL JDBJECTxurrent_scenario = SCENARIO_94 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = HGM_9 

# New best scenario 

# CONTROL J5BJECTxurrent_scenario = SCENARIO_109 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = DIFFUSER_7 

# New best scenario 

# CONTROL JDBJECTxurrent_scenario = SCENARIO_116 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = CCVJ 

# New best scenario 

# CONTROL J)BJECTxurrent_scenario = SCENARIO_120 

# Next behavior to expand 

# CONTROL _OBJECT2.is_now = MKERJO 

# New best scenario 

# CONTROL JDBJECT.current_scenario = SCENARIO_133 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = NZL_COOL JO 

# New best scenario 

# CONTROL JDBJECTxurrent_scenario = SCENARIOJ34 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = MFV_7 

# New best scenario - 

# CONTROL JDBJECTxurrent_scenario = SCENARIO_135 

# Next behavior to expand 

# CONTROL JDBJECT2.is_now = LPFPJ3 

# New best scenario 

# CONTROL JDBJECTxurrent_scenario = SCENARIO_146 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPFT_3 

# New best scenario 

# CONTROL JDBJECTxurrent_scenario = SCENARIO_155 

# Next behavior to expand 

# CONTROL jDBJECT2.is_now = F109_6 

# New best scenario 

# CONTROL JDBJECT.currentjscenario = SCENARIO_169 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = MCCCOOL9 

# New best scenario 

# CONTROL^OBJECTxurrent^scenario = SCENARIO_144 

# Fault assumption looks wrong: increase estimated cost 

# New best scenario 

# CONTROL JDBJECTxurrent_scenario = SCENARIO J74 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPOT_78 

# New best scenario 

# CONTROL jDBJECTxurrent_scenario = SCENARIO J97 

# No fault! 

# Next behavior to expand 

# CONTROL J>BJECT2.is_now = O203_10 

# New best scenario 

# CONTROL_OBJECTxurrent_scenario = SCENARIOJ98 

# Next behavior to expand 



32 g Final Report, July 1993 



— # CONTROL JDBJECT2.is_now = HPFTJL1 

# New best scenario 

i i # CONTROL JDBJECLcurrent_scenario = SCENARIO JL99 

:_: # Next behavior to expand 

# CONTROL_OBJECT2.is_now = HGM_23 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_214 
~~~ # Next behavior to expand 

# CONTROL_OBJECT2.is_now = DIFFUSER_7 
^ ; # New best scenario 

-~ # CONTROL_OBJECTcurrent_scenario = SCENARIO_221 

# Next behavior to expand 

# CONTROL JDBJECT2.is_now = CCV Jl 

# New best scenario 

~" # CONTROL_OBJECT.current_scenario = SCENARIO_225 

-^ # Next behavior to expand 

r i # CONTROL_OBJECT2.is_now = MKER_44 

** # New best scenario 

# CONTROL JDBJECT.currcnt_scenario = SCENARIO_238 
" : # Next behavior to expand 

~ # CONTROL_OBJECT2.is_.now = NZL_COOL _17 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_239 

# Next behavior to expand 

— # CONTROL JDBJECT2.is_now = MFV_21 

# New best scenario 

— # CONTROL_OBJECT.cnrrent_scenario = SCENARIO_240 

— # Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPFP_3 

# New best scenario 

„ # CONTROL_OBJECT.current_scenario = SCENARIO_253 

# One fault in behavior 

__ # CONTROL JDBJECT3.bestJaultyJbehavior = LPFP_17 

-™ # Fault type 

— # C0NTR0L_0BJECT3.best_current Jault = LOWJEFFICENCY 

# Next behavior to expand 

lg # C01NnHOLJ3BJECT2.is_now = LPFT_24 

S3 # New best scenario 

# CONTROL JDBJECTxurTent_scenario = SCENARIO_257 

# Next behavior to expand 

ZZ # CONTROL_OBJECT2.is_now = F109_17 

# New best scenario 

# CONTROL_OBJECTxurrent_scenario = SCENARIO_265 
__" # Next behavior to expand 

^ # CONTROL_OBJECT2.is_now = MCC_COOL_19 

# New best scenario 

g # CONTROL JDBJECTxurrent_scenario = SCENARIO_252 

|g # One fault in behavior 

# CONTROL_OBJECT3.best_faulty_behavior = LPFP_16 
s # Fault type 

g # C0NTR0L_0BJECT3.best^current_fault = LOWJEFFICIENCY 

** # Fault assumption looks wrong: increase estimated cost 

# New best scenario 

Ei # CONTROL OBJECT.current scenario = SCENARIO 252 



Final Report, July 1993 329 



# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPFT_36 

# New best sce nari o 

# C0NTR0L_OBJECT.current_scenario = SCENARIO_268 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = F109_27 

# New best scenario 

# CONTROL J3BJECTxurrent_scenario = SCENARIO_276 

# Next behavior to expand 

# CONTROL^OBJECm.is^now = MCC_COOL_28 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_269 

# Fault assumption looks wrong: increase estimated cost 

# New best scenario 

# CONTROL JDBJECTxunentj>cenario = SCENARIO_251 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPFT_45 

# New best scenario 

# CONTROL jDBJECTxurrent_scenario = SCENARIO_285 

# No fault! 

# Next behavior to expand 

# CONTROL_OBJECT2as_now = F109J37 

# New best scenario 

# CONTROLjOBJECT.current_scenario = SCENARIO_299 

# Next behavior to expand 

# CONTROL J)BJECT2.is_now = MCC.COOL J>1 

# New best scenario 

# CONTO0LjDBJECTxurrent_scenario = SCENARIO_300 

# One fault in behavior 

# CONTROL_OBJECT3.best JauItyJ>ehavior = MCC_COOL_43 

# Fault type 

# CONTROL J3BJECT3.best_currentJault = LEAK 

# Next behavior to expand 

# CONTROL jDBJECT2.is.jiow = FPB_6 

# New best scenario 

# CONTROL J}BJECTxurrent_scenario = SCENARIO_309 

# Next behavior to expand 

# CONTROLjDBJECT2.is_now = FPOV JX 

# New best scenario 

# CONTROL JDBJECTxurrent_scenario = SCENARIO_317 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = O206J7 

# New best scenario 

# CONTROL_OBJECTxurrent_scenario = SCENARIO_318 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = F107 JO 

# New best scenario 

# CONTROL JDBJECTxurrent_scenario = SCENARIO_324 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = M103_14 

# New best scenario 

# CONTROL_OBJECTxurrent_scenario = SCENARIO_327 

# Next behavior to expand 

# CONTROL OBJECT2.is now = F110 9 



330 Fina l Report, July 1993 



# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_328 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = F108_8 

# New best scenario 

# CONTROL OBJECT.current scenario = SCENARIO 329 



Hi 



Final Report, July 1993 331 



Execution Transcript: Using PBM and Heuristic Rules 



A.5.6MCC Leak Example: Execution transcript using PBM and heuristic rules 

Execution Transcript: Using PBM and Heuristic Rules 

NEXPEKT Serial Number 1-2.0B-S4X1-051091-1458 
Copy for Client of Neuron Data, Inc. 

# . NEXPERT. - Copyright (C) 1986 - 1990 by NEURON DATA. Copyright is claimed in both the underlying comput- 
er program and the resulting output in the form of an audiovisual work. 

# Customer or User is not permitted to make any copies of this software (NEXPEKT) for any purpose. This software 
is a confidential trade secret of NEURON DATA Inc. Refer to the license agreement. 

# Next behavior to expand 

# CONTOOL J)BJECT24S jiow = MCC_COOL_4 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_20 

# One fault in behavior 

# CONTROL_OBJECT3.best_faulty_behavior = MCCCOOLJ76 

# Fault type 

# CONTROL_OBJECT3.best_cunent_fault = LEAK 

# Next behavior to expand 

# CONTROL JDBJECT2.is_now = DIFFUSER_11 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_44 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = CCV_6 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_46 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = FUEL_FLOW_CTRL_2 

# New best scenario 

# CONTROL jOBJECT.current_scenario = SCENARIO_47 

# Next behavior to expand 

# CONTROL J3BJECT2.is_now = F101_4 

# New best scenario 

# CONTROL_OBJECr.current_scenario = SCENARIO_57 

# Ignoring pin deviation at HIPUMP 

# Next behavior to expand 

# CONTROL JDBJECT2.is_now = HPFP_4 

# New best scenario 

# CONTROL_OBJECTxurrent_scenario = SCENARIO_59 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = HPFTJ5 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_60 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = HGM_5 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_101 

# Next behavior to expand 

# CONTROL jmrECT2.is_now = HPOT_6 

# New best scenario 

# CONTROL_OBJECTxurrent_scenario = SCENARIOJ02 

# Next behavior to expand 

# CONTROL JDBJECT2.is_now = HPOP_PBP_6 

# New best scenario 

# CONTROL OBJECT.current scenario = SCENARIO 107 



332 Final Report, July 1993 



w # Next behavior to expand 

# CONTROL J3BJECT2.is_now = M104_5 
'— # New best scenario 

# CONTROL J3BJECT.current_scenario = SCENARIO_120 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = O205_5 

# New best scenario 

~ # CONTROL J3BJECT.current_scenario = SCENARIO_126 

# Next behavior to expand 

# CONTROL J3BJECT2.is_now = M101_8 

— # New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_129 
_ # Next behavior to expand 

rl # CONTROL_OBJECT2.is_now = OFOVJ7 

# New best scenario 

# CONTROL JDBJECT.cunent_scenario = SCENARIO_131 

# Next behavior to expand 

— # CONTROL_OBJECT2.is_now = OPB_9 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_137 
~ # Next behavior to expand 

# CONTROL_OBJECT2.is_now = F108J7 

# New best scenario 

H # CONTROL_OBJECT.current_scenario = SCENARIO_143 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = M103_9 
^ # New best scenario 

— # CONTROL JDBJECT.current_scenario = SCENARIO_146 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = ¥110 J 
jj # New best scenario 

# CONTROL_OBJECT.cunent_scenario = SCENARIO_148 

# Next behavior to expand 

B # CONTROL_OBJECT2.is_now = FPB_7 

** # New best scenario 

# CONTROL JDBJECT.current_scenario = SCENARIO_150 
p^ # Next behavior to expand 

II #CONITlOL_OBJECT2.is M now = FPOV_ll 

# New best scenario 

^ # CONTROL JDBreCT.current_scenario = SCENARIO_158 

g # Next behavior to expand 

"" # CONTROL_OBJECT2.is_now = O206J7 

# New best scenario 

lj # CONTROL J}BJECT.current_scenario = SCENARIO_159 

fe» # Next behavior to expand 

# CONTROLJDBJECT2.is_now = NOZZLEl_9 
^ # New best scenario 

§j # CONTROL_OBJECT.current_scenario = SCENARIOJL60 

# Next behavior to expand 

~ # CONTROL_OBJECT2.is_now = MCC_10 

§j # New best scenario 

" # CONTROL j3BJECTxurrent_scenario = SCENARIO_163 

# Next behavior to expand 

# CONTROL OBJECT2.is now = MOV 9 



Final Report, July 1993 333 



# New best scenario 
#CONTROL_OBJECT.current -> scenario = SCENARIO_171 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = O204_7 

# New best scenario 

# CONTROL JDBJECTxurrent_scenario = SCENARIO_172 

# Next behavior to expand 

# CONTROL_OBJECT2.is.now = F107_9 

# New best scenario 

# CONTROL_OBJECTxunent ^scenario = SCENARIO_174 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = MKERJ4 

# New best scenario 

# CONTROL JDBJECTxurrent_scenario = SCENARIOJ77 

# Next behavior to expand 

# CONTROL J)BJECT2.isjiow = NZL_COOL_12 

# New best scenario 

# CONTROL_OBJECTxurrent_scenario = SCENARIO_178 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = MFVJ7 

# New best scenario 

# CONTROL jDBJECTxurrent_scenario = SCENARIO_179 

# Next behavior to expand 

# CONTROL_OBJECT2,is_now = F109_5 

# New best scenario 

# CONTROL JDBJECTxurrent_scenario = SCENARIO_188 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPFT_4 

# New best scenario 

# CONTROL JDBJECTxurrent_scenario = SCENARIO_193 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LPFP_5 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARIO_201 

# Next behavior to expand 

# CONTROL_OBJECT2.is jiow = O203_6 

# New best scenario 

# CONTROL_OBJECt.current_scenario = SCENARIO_207 

# Next behavior to expand 

# CONTROL_OBJECT2.is_now = LP0T_4 

# New best scenario 

# CONTROL_OBJECT.current_scenario = SCENARI0_213 

# Next behavior to expand 

# C0OTR0L_0BJECT2.is jiow = LP0P_7 

# New best scenario 

# CONTROL J3BJECT.current_scenario = SCENARIO_215 

# Next behavior to expand 

# CONTROL J5BJECT2.is_now = O201_3 

# New best scenario 

# CONTROL OBJECTxurrent scenario = SCENARIO 218 



334 



Final Report, July 1993